broadband in asia€¦ · 0.1 the rise of broadband internet has created opportunities for...

Broadband in Asia–Pacific: How investment, partnerships and policy are driving a global success story | i

Ref: 2004569-383 Error! U nknown document propert y name.

Contents

0 Executive summary 1

0.1 The rise of broadband Internet has created opportunities for Asia-Pacific telecoms operators

to thrive 1

0.2 CAPs are helping to drive growth and opportunities for IAPs and other participants in the

Internet value chain 6

0.3 Partnerships and a forward-looking policy environment are essential to unlocking further

the benefits of the Internet 13

1 About this report 18

2 The rise of broadband Internet has created opportunities for Asia–Pacific telecoms

operators to thrive 21

2.1 Broadband Asia-Pacific: a strong growth story, but wide variations remain 21

2.2 The current picture: broadband is replacing legacy services as the main revenue and profit

driver for IAPs 23

2.3 Investing in the future: IAPs in Asia–Pacific are increasingly leveraging their customer

relationships to drive adoption of innovative OTT services 28

3 CAPs are helping to drive growth and opportunities for IAPs and other participants

in the Internet value chain 32

3.1 Consumers take up broadband services explicitly to access content and services created by

CAPs 32

3.2 Quality of service drives take-up and usage, and CAPs are investing alongside IAPs in

network infrastructure to improve the user experience of the Internet 40

3.3 Challenges remain in bringing the benefits of broadband to more people and to

‘unconnected’ countries 44

4 Partnerships and a forward-looking policy environment are essential for unlocking

further the benefits of the Internet 49

4.1 CAPs and IAPs are forming innovative partnerships to address these barriers 49

4.2 Some policy makers are implementing innovative policy initiatives to lower barriers, often

in partnership with IAPs or CAPs 54

4.3 Policy makers have an essential role to play in ensuring that barriers to innovation and

investment are as low as possible 58

5 Conclusion 65

Annex A Primary research programme

Annex B State of the broadband ecosystem in Asia–Pacific

Annex C Key barriers to broadband development in Asia–Pacific

ii | Broadband in Asia–Pacific: How investment, partnerships and policy are driving a global success story

Ref: 2004569-383

Confidentiality Notice: This document and the information contained herein are strictly private

and confidential, and are solely for the use of Google.

Copyright © 2015. The information contained herein is the property of Analysys Mason

Limited and is provided on condition that it will not be reproduced, copied, lent or

disclosed, directly or indirectly, nor used for any purpose other than that for which it was

specifically furnished.

Analysys Mason Limited

Bush House, North West Wing

Aldwych

London WC2B 4PJ

UK

Tel: +44 (0)20 7395 9000

Fax: +44 (0)20 7395 9001

[email protected]

www.analysysmason.com

Registered in England No. 5177472

This report was commissioned and sponsored by Google, and prepared independently by

Analysys Mason, a global consultancy specialising in telecoms, media and technology.

The analysis contained in this document is the sole responsibility of Analysys Mason and

does not necessarily reflect the views of Google or other contributors to the research.

We would like to thank our research partners, On Device Research and 2-morrow, as well as

Alex Kozak, Aparna Sridhar, Austin Schlick, Breanna Swart, Derek Slater, Goh Lih Shiun

and Robindhra Mangtani from Google for their valuable comments.

Broadband in Asia–Pacific: How investment, partnerships and policy are driving a global success story | 1


0 Executive summary

This report explores how the broadband ecosystem has developed in the Asia-Pacific region1 over

the past 15 years, and the opportunities that this has created for consumers and telecoms operators.

Our research finds that:

The growth of the broadband ecosystem has created opportunities for telecoms operators to

thrive as Internet Access Providers (IAPs), and to develop new content and applications

themselves;

Content and application providers (CAPs) are helping to create growth opportunities for IAPs

and other market participants; and

A conducive policy environment is essential to the further growth of the Internet.

0.1 The rise of broadband Internet has created opportunities for Asia-Pacific telecoms

operators to thrive

Internet markets in the region have seen strong growth. Take-up of broadband services has grown

at 28% per annum since 2005, while usage of online applications and content has grown rapidly, at

28% per annum since 2012. Despite this strong overall growth, the level of development of the

region’s broadband markets varies widely. To reflect this, we have divided the 21 countries of

interest into three ‘clusters’:

Connected countries, which enjoy world-class connectivity, with near-ubiquitous coverage of

3G and 4G mobile and fixed broadband, high take-up, and high usage of advanced services.

This includes Australia, New Zealand, Singapore, Hong Kong, Japan, Korea, and Taiwan.

Connecting countries, which have high coverage, but moderate take-up of 3G and 4G mobile

broadband, low coverage and take-up of fixed broadband, and increasing usage of advanced

Internet services. This includes Malaysia, Thailand, Philippines, Indonesia, Brunei, Vietnam,

Cambodia, and Sri Lanka.

Unconnected countries, which have yet to develop advanced broadband services. They have

very little coverage or take-up of either 3G mobile services or advanced fixed broadband

services. This includes Papua New Guinea, Bangladesh, Pakistan, Nepal, Myanmar, and Laos.

These clusters are illustrated in Figure 0.1.

1 We include the following 21 countries in our analysis of the Asia-Pacific region: Australia, Bangladesh, Brunei

Darussalam, Cambodia, Hong Kong, Indonesia, Japan, Laos, Malaysia, Nepal, New Zealand, Pakistan, Papua New Guinea, the Philippines, Singapore, South Korea, Sri Lanka, Taiwan, Thailand and Vietnam. India and China are not covered in this report, their scale and diversity warranting separate reviews of their own.

2 | Broadband in Asia–Pacific: How investment, partnerships and policy are driving a global success story

Ref: 2004569-383

Figure 0.1: Illustrative map of countries in each cluster [Source: Analysys Mason, 2015]

0.1.1 The current picture: broadband is replacing legacy services as the main revenue and profit

driver for IAPs

This strong growth in broadband has resulted in growth in overall telco revenues since the

introduction of advanced broadband. Figure 0.2 shows this revenue growth for all services since

the introduction of advanced broadband.2

2 Note that we exclude the ‘unconnected’ countries, as these have yet to develop advanced broadband.

Connected

Connecting

Unconnected


Ref: 2004569-383

Figure 0.2: Overall IAP

revenue, indexed to the

year of 3G launch,

‘connected’ and

‘connecting’ clusters,

using constant

exchange rates

[Source: GSMA,

company data, ITU,

TeleGeography,

Analysys Mason, 2015]

In addition, the importance of data and Internet services within the IAP revenue mix has increased

strongly. This is illustrated in Figure 0.3, from the year of 3G launch. ‘Connecting’ countries have

rapidly caught up with ‘connected’ countries, demonstrating the importance of the Internet to the

telecoms sectors of middle-income countries.

Figure 0.3: Proportion

of total service revenue

from data and internet

services [Source:

GSMA, company data,

ITU, TeleGeography,


0

50

100

150

200

250

300

350

200

1

2002

200

3

200

4

200

5

2006

200

7

200

8

200

9

201

0

201

1

201

2

201

3

201

4

Re

ve

nu

e in

de

xe

d t

o 1

00

Connected Connecting

Connected 3G launch Connecting 3G launch

0%

10%

20%

30%

40%

50%

60%

70%

0 1 2 3 4 5 6 7 8 9 10 11 12 13

Pro

po

rtio

n o

f re

ve

nu

e fro

m d

ata

Years since launch of 3G



Ref: 2004569-383

As revenue has grown and shifted from voice to data services, the source of IAPs’ profitability3

has also shifted, from legacy telephony services to new data services, including broadband Internet

access. Over the period since the launch of 3G (broadly the last 10-15 years, depending on the

market), total EBITDA has been constrained, but relatively stable in all countries in the region

(reflecting many factors such as subsidies for high-end handsets).

With modern networks, IAPs are able to provide all their services with the same infrastructure.

The shift in revenue towards data services, together with the low incremental cost of supplying

them, implies that data services are now also a significant (and increasing) source of IAP profits.

Key to this are developments in 4G networks, which are fundamentally data networks that allow

spectrum and network resources to be shared seamlessly between services. Likewise the

development of converged core networks allows all voice and data traffic to be carried seamlessly

over an IP network. As a result, the profits made by IAPs are now mainly attributable to data

services, a pattern that is set to continue for years to come.

Investments made in these modern networks have produced healthy returns. Overall industry

returns, measured by the Return on Capital Employed (ROCE) ratio,4 are illustrated below. These

appear healthy overall, comparable to estimates of the cost of capital for telecoms operators

(shown as the green band in Figure 0.4)5 for mature markets and at a lower level for countries

where overall expected economic growth is much faster, network upgrades are recent or under-

way, and where equipment costs tend to be larger compared to their revenues.

Figure 0.4: ROCE for

IAPs in ‘connected’ and

‘connecting’ clusters

[Source: Analysys

Mason, company data,

GSMA, TeleGeography,

ITU, Aswath

Damodaran (NYU

Stern), 2015]

* The ROCE for Thailand is for mobile only.

3 We use earnings before interest, taxes, depreciation and amortisation, or EBITDA, as a proxy here

4 ROCE is defined as earnings before interest and tax divided by total capital employed

5 Compiled based on data by Aswath Damodaran, NYU Stern, 2015; the range shown is for telecom services and

wireless providers, and comprises Global, Emerging markets and Japan values

0%

5%

10%

15%

20%

25%

Au

str

alia

Hon

g K

on

g

Ne

w Z

ea

land

Sin

ga

pore

Ta

iwa

n

Ma

laysia

Jap

an

Sri

La

nka

Th

aila

nd

*

So

uth

Ko

rea

Vie

tna

m

Indo

nesia

Ph

ilipp

ines

RO

CE



Ref: 2004569-383

0.1.2 Investing in the future: IAPs in Asia-Pacific are increasingly leveraging their customer

relationships to drive adoption of innovative OTT services

In addition to these returns, IAPs have begun to invest in new OTT services. Investments have

been made by IAPs in developing and developed countries, in a range of verticals. Whilst these

initiatives are in their early stages, there is evidence of IAPs in the region achieving growth in

these areas.

Examples of these investments in developing countries include:

Singtel’s investments the Amobee advertising platform and the Hooq OTT video service;

PLDT’s investments in its Smart Money payment app which allows overseas workers to send

remittances back home, and its Smart Pinoy app which allows Filipinos overseas to send gifts

to loved ones at home;

Grameenphone’s investment in ‘WowBox’, a zero-rated mobile content app;

Developed country examples include:

NTT DoCoMo’s investments in the ‘dmarket’ mobile application portal, which provides

Japan-specific applications and content;

KDDI’s Smart Pass service, which allows access to the top 500 Android apps for a fee; and

Singtel’s partnership with Fetch TV, a subscription-based OTT television service, and its TV

GO service in Singapore, which provides access to pay TV content to Singtel’s existing TV

subscribers.

In developing countries, IAPs are developing services that are tailored to specific audiences, in a

way that is not easily doable for global CAPs. For example, PLDT is targeting overseas Filipino

workers, and Grameenphone targets price sensitive prepaid customers. IAPs do this by taking

advantage of their existing customer relationships and knowledge to provide a tailored service,

seamless billing, and targeted advertising.

Consumers in developed markets already have access to advanced online content. IAPs therefore

are using their customer relationships, sales channels, billing relationships, and networks to act as

distributors of advanced content. They either do this as partners with CAPs (e.g. KDDI’s Smart

Pass, or Optus with Fetch TV) or over their own platforms (e.g. NTT DoCoMo’s dmarket,

Singtel’s TV GO service).

These investments are at an early stage, but the fact that IAPs are committing significant sums to

them speaks to the opportunities created by advanced broadband. As the Internet reaches more

people in in the region, these opportunities are likely to increase further, particularly where IAPs

have a unique position in understanding their customers and processing payments.


Ref: 2004569-383

0.2 CAPs are helping to drive growth and opportunities for IAPs and other participants

in the Internet value chain

0.2.1 Consumers take-up broadband services explicitly to access content and services from CAPs

For Internet adopters, online content and applications are the main drivers behind usage, the

decision to upgrade their access and the value they get from the Internet

CAPs create content and services that cause consumers to sign up for broadband access. This is

well documented. For example, research performed in 2013 for the Computer and

Communications Industry Association (CCIA) found that European consumers using advanced

services are more likely to upgrade their connection. 40% of those making video calls and 49% of

those uploading video upgraded in the year preceding the study6.

Likewise, new research commissioned for this study7 found that there is strong latent demand for

ever better services, linked to the usage of advanced content. We found that between 29% and

49% of fixed users and between 25% and 50% of mobile users responded that they would upgrade

in order to receive a better user experience. In addition, a majority of current Internet users who

stated that they would upgrade their fixed and mobile services (‘upgraders’) are significantly more

likely to use advanced services such as video:

69% of mobile upgraders use video several times a week or more, compared to 57% of non-

upgraders.

82% of fixed upgraders use video regularly, compared to 76% of non-upgraders.

The responses of fixed and mobile upgraders is shown in Figure 0.5 and Figure 0.6

6 http://www.ccianet.org/2013/09/new-research-shows-high-bandwidth-online-services-create-demand-for-faster-

more-expensive-broadband/

7 Analysys Mason has commissioned primary research of Internet adopters in Australia, Indonesia, Singapore, Sri

Lanka, and Thailand and non-adopters in Indonesia, Sri Lanka, and Thailand. For more details of the methodology and findings of this research, please refer to Annex A.


Ref: 2004569-383

Figure 0.5: How often do you use your fixed Internet

connection for the following services, asked of fixed

upgraders8

(n=1696) [Source: Analysys Mason,

2015]

Figure 0.6: How often do you use your mobile

Internet connection for the following services, asked

of mobile upgraders9 (n=2894) [Source: Analysys

Mason, 2015]

Our research examined users’ willingness to pay for Internet services, relative to what they

currently pay (Figure 0.7). For fixed services, upgraders have a higher willingness to pay than non-

upgraders, although this relationship is weaker for mobile users. This suggests that, at least for

fixed users, the drive to upgrade, itself linked to usage of advanced services, is linked to a higher

willingness to pay.

The research also shows that there appears to be significant untapped value in ‘connecting’

countries (here Thailand, Sri Lanka, and Indonesia). As shown in Figure 0.7, the willingness to

pay of upgraders in these countries has a greater premium over non-upgraders when compared to

‘connected’ countries (Australian and Singapore). This suggests that if these countries could drive

upgrades, they could unlock this high ‘willingness to pay’ to the benefit of both consumers and

IAPs. We discuss initiatives and policies that can facilitate this in Section 0.3, notably by lowering

barriers to network investments to expand coverage of faster networks and improve quality of

service.

8 Fixed upgraders are defined as mobile users who gave a response other than “I don’t want to upgrade”, “I am

already receiving the best connectivity available at my location”, “Other”, or “Don’t know” to the question “What would make you consider upgrading the current fixed Internet service?”

9 Mobile upgraders are defined as mobile users who gave a response other than “I don’t want to upgrade”, “I am

already receiving the best connectivity available at my location”, “Other”, or “Don’t know” to the question “What would make you consider upgrading the current mobile Internet service?”

0% 50% 100%

Listening to music

Playing games

Watching videos

Communications

Social networking

Browsing

Every day Several times a week

Once a week Once a month

Very rarely Never

0% 50% 100%

Listening to music

Watching videos

Communications

Playing games

Browsing

Social networking



Very rarely Never


Ref: 2004569-383

Figure 0.7: Fixed user

willingness-to-pay

premium above current

price paid (Australia

n=404, Indonesia

n=297, Singapore

n=520, Sri Lanka

n=137, Thailand n=376)

[Source: Analysys

Mason, 2015]

Creating and raising awareness of content and applications relevant and accessible to non-

adopters is the most important factor to improve Internet take-up

As part of our primary research programme, we spoke to ‘non-adopters’10

in Indonesia, Sri Lanka

and Thailand to discuss their motivations and understand what they perceive as the main reasons

not to adopt or use the Internet (Figure 0.8).

The most important reason cited was “awareness”,11

mentioned by 75% of respondents. This is a

measure of whether consumers are aware of the benefits of consuming Internet content. Service

availability (driven by access to a broadband network) was less important. Literacy, the lack of a

connected device, and the availability of relevant content appear to be barriers for a significant

minority of people.

These results are also borne out by other research. Research conducted in Singapore in 2014 by the

national regulatory authority, IDA12

, found that, of non-adopters, 55% said they did not need an

Internet connection, and 30% said they did not know how to use the Internet, showing that

awareness is also a problem for non-adopters in Singapore.

10 Defined as people who do not have either a personal mobile Internet service or a fixed Internet service at their place

of residence

11 Defined as respondents who answered “I don’t trust it / don’t think it’s safe”, “I don’t need it”, “I don’t know what the

Internet is or what it does”.

12 IDA Consumer Awareness and Satisfaction Survey, 2014

0% 40% 80% 120%

Australia

Indonesia

Singapore

Sri Lanka

Thailand

Fixed Non-upgraders Fixed Upgraders


Ref: 2004569-383

Figure 0.8:Reasons cited for non-adoption or non-usage of Internet services as stated by the interviewees in

all countries [Source: Analysys Mason, 2015]

Notably, the majority of non-adopters have never even used the Internet before, at, for example, an

Internet café, home of a friend or family member, or at work or an educational institution. This is

the case for 66% of respondents in Indonesia, 82% in Sri Lanka, and 76% in Thailand. This lack of

experience of the Internet is also likely to drive lack of awareness of its benefits.

In this context, future growth in adoption of Internet access services by current ‘non-adopters’

could be driven by awareness-raising initiatives and investments in content that is relevant and

accessible to these new users. CAPs, potentially including some IAPs, will play a central role in

this evolution.

0.2.2 Quality of service drives take-up and usage, and CAPs are investing alongside IAPs in

network infrastructure that improves user experience of the Internet

Quality of service is linked to service upgrades for both fixed and mobile users

The quality of service received by an end user helps to drive users to upgrade their services.

Indeed, users are more likely to upgrade their service if they have had a good experience of

consuming advanced content. In our research, we asked respondents about the quality of their

experience when watching video content. Both fixed and mobile upgraders were more likely than

non-upgraders to be able to watch video with no degradation in quality at least most of the time.

83% of fixed upgraders received good quality video at least most of the time compared to 76% of

fixed non-upgraders, with 69% of mobile upgraders against 56% of mobile non-upgraders

receiving good video services most of the time.

0% 10% 20% 30% 40% 50% 60% 70% 80%

Other

It's too expensive

Someone close to me can use it for me when Ineed something

Availability

I don't own a connected device

Literacy

Relevance of content

Awareness


Ref: 2004569-383

CAPs are increasingly driving the quality of user experience that stimulates adoption, usage and

upgrades of Internet access services

CAPs increasingly make investments in networks and infrastructure. This is in addition to the

substantial investments IAPs are making in upgrading their core and access networks. CAPs are

investing in many aspects of networks, from hosting to transport to the delivery of traffic to IAPs

as close as possible to end-users.

A large part of the investment made by CAPs is in hosting infrastructure, including data centres,

servers and storage devices. This hosts and processes the content and applications developed by

CAPs, and houses nodes where content and applications are distributed and cached. Example

Asia-Pacific investments include Google’s data centres in Singapore and Taiwan, and Amazon

Web Services (AWS) data centres in Singapore, Sydney, Tokyo and Beijing.

CAPs also invest in transport networks. These are long-distance submarine or terrestrial fibre-

optic cables that carry Internet traffic around the world. Whilst IAPs and backbone operators are

large investors in these, CAPs are now also major contributors. Example CAP investments in

Asia-Pacific include APX-East (between Australia and the USA), FASTER (between Japan and

the USA) and BBG (between South-East Asia, the Middle East and Europe).

Finally, CAPs invest in delivery infrastructure, which carries traffic from long-distance transport

networks to IAP networks. This includes the points of interconnect where networks exchange

traffic (either at Internet exchanges or private points of interconnect). It also includes content

delivery networks (CDNs) that help to optimise traffic delivery, for example by shifting content to

hosting infrastructure that is closer to end users.

These investments partly reduce the requirement for IAPs to invest themselves, for example to

provide required network capacity to carry Internet traffic. Their main impact on IAPs is to reduce

capital and operating costs.

They are also partly incremental. Incremental investments improve the quality of service that end

users experience when consuming content and applications. CAPs make these investments to

attract more users to their services, but the investments also benefit consumers and IAPs.

In a 2014 paper,13

Analysys Mason estimated the magnitude of the global investments by CAPs in

networks and infrastructure. We estimated global CAP investments to be USD33 billion per

annum, of which USD7.7 billion was in Asia–Pacific.14

Investments in submarine and terrestrial

cable infrastructure were estimated to average USD2.7 billion per annum, of which

USD700 million was in Asia–Pacific.

13 See http://www.analysysmason.com/Research/Content/Reports/Content-application-provider-Internet-infrastructure-

Sept2014/.

14 This figure includes the entire Asia-Pacific region


Ref: 2004569-383

0.2.3 Challenges remain in bringing the benefits of broadband to more people and to

‘unconnected’ countries

We saw that a lack of awareness of the benefits of the Internet is one of the foremost drivers of

non-adoption and non-usage. This creates risks associated with potential demand, which can deter

investment in infrastructure, especially in ‘unconnected’ countries. IAPs may find it too risky to

invest in new network coverage, given apparent demand for Internet services is low. Here we

explore these other barriers, and the impact they have had on the ‘unconnected’ countries.

There are a wide range of barriers that can have an impact on broadband development outcomes

We have identified a wider range of barriers that can influence all aspects of broadband

development, especially in the ‘unconnected’ countries. We can define drivers and barriers to

development with reference to the outcome that they influence:

The availability of broadband services determines whether end users are able to use these

services at all. It includes the coverage access networks, backbone, hosting and other

infrastructure

The take-up of services reflects the extent to which consumers decide to become Internet

users and sign up for services from fixed and mobile operators

The usage of services refers to how much consumers actually use their Internet connections to

access content and applications.

For each of these outcomes, we have identified a set of potential drivers (if they are present) or

barriers (if they are lacking). An overview of these is set out in Figure 0.9.

Figure 0.9: Drivers and barriers related to availability, take-up and usage [Source: Analysys Mason, 2015]

Driver /

barrier

Description

Drivers and barriers related to availability

Business

case

Is there an economic business case for rolling out broadband coverage somewhere new?

Is there a policy case for it and the money to fund it? Roll-out requires fixed and mobile

access networks, long-distance domestic and international backbone networks, and

hosting and interconnection infrastructure.

Two key factors play a role here: the cost of inputs such as land, equipment, spectrum

and licences; and the risk associated with demand from consumers.

Availability

of inputs

Are the key inputs required available at all? Scarcity of inputs affects three areas:

Radio spectrum and land are important for mobile networks (handled by governments

and regulators)

Reliable electricity supply is important for networks and for users to recharge their

devices

Existing networks that can be shared, including domestic networks and international

connectivity, are important in minimising roll-out costs.

Policy and In addition to competition and spectrum management, this includes economic regulation


Ref: 2004569-383

Driver /

barrier

Description

regulation of telecoms networks and services. Obligations attached to telecoms licences are

important drivers of roll-out. Also, policy makers increasingly include national broadband

plans as a key policy plank to spur wider economic development. Finally, innovation in

network roll-out is beginning to involve other areas of regulation. For example, Google’s

Project Loon (see Section 4) is increasing the relevance of aeronautical regulation.

Drivers and barriers related to take-up

Cost and

device

availability

Does the cost to the consumer of a fixed or mobile Internet service make it unaffordable?

Is the cost and payment schedule of a device a barrier to affordability? Are attractive

devices available in the market?

These aspects can be influenced by the market structure and strategy of device

manufacturers, whether IAPs subsidise devices, and whether there are restrictive

government policies or taxes (including import restrictions or duties).

Social

factors

Are consumers able to use the Internet, driven by their education and awareness or by

cultural restrictions? This include awareness of the benefits of Internet usage (which is

influenced by both literacy and IT literacy); and whether the use of the Internet is deemed

appropriate for all users in certain communities.

Drivers and barriers related to usage

Means of

access

How do consumers access the Internet? Is it on a PC at home or work, on a smartphone

or tablet, or in an Internet café? This affects the content and applications that a consumer

uses: for example, long videos are easier to consume at home.

Relevant,

accessible

content

Does interesting content exist in the local language, made by local or global CAPs?

Quality of

service

Can the consumer receive content at sufficiently high bandwidths and with low enough

latency? This is important for advanced content.

Trust What is the perceived and actual level of security, driven by actual security and

consumer awareness of security. Is Internet access broadly free and private? Monitoring

and censorship can make consumers feel uncomfortable about accessing content freely.

‘Unconnected’ countries have poorer outcomes due to the greater impact of key barriers

‘Unconnected’ countries perform more poorly than the other clusters on all three outcomes

discussed above. They have lower availability, with an average 3G coverage of 60%, and lower

take-up, with a take-up of 3G and 4G services of just 3%15

. Moreover, users in ‘unconnected’

countries who do take up Internet services also have substantially lower usage than users in other

clusters: 17GB per annum per subscriber, compared to 157GB16

per subscriber in connected

countries.

15 Source: GSMA

16 Source: Analysys Mason


Ref: 2004569-383

These poorer outcomes are in part driven by the greater impact of key barriers. We saw in above

that a lack of awareness is a key barrier to take-up, and this is also the case in ‘unconnected’

countries. This barrier affects the business case for network roll-out as well. In addition, a number

of other barriers apply to a greater extent in ‘unconnected’ countries than in other clusters:

The cost of service plans and handsets acts as a barrier to take-up

Low PC penetration affects the types of content that users can access at a reasonable quality

There is less relevant content available in local languages.

CAPs and IAPs are aware of these barriers that exist in the ‘unconnected’ cluster. They are

currently implementing an array of partnerships and programmes to address them. We explore

these more fully in the next section.

0.3 Partnerships and a forward-looking policy environment are essential to unlocking

further the benefits of the Internet

Here we explore how the barriers to further broadband growth in Asia–Pacific can be reduced or

removed.

CAPs are becoming essential partners in the development of broadband Internet opportunities.

They are forming partnerships with IAPs and other organisations to help lower the barriers to

Internet adoption and create new opportunities for mutual revenue growth.

Proactive policy makers are implementing programmes that help to lower barriers to adoption,

sometimes in partnership with IAPs or CAPs.

Policy makers and regulators also play an essential role by ensuring that the environment they

influence is as conducive as possible to investment and innovation, by IAPs, CAPs and others.

We discuss each of these in turn.

0.3.1 CAPs and IAPs are forming innovative partnerships to address these barriers

CAPs and IAPs are increasingly working together and with other industry partners to their mutual

benefit to address barriers to Internet availability, take-up and usage. An overview of a number of

example partnerships covered is provided in Figure 0.10, setting out the key barrier addressed, and

a summary of the outcome.

Figure 0.10: Overview of example CAP partnerships [Source: Analysys Mason, 2015]

CAP Overview Barrier

addressed

Outcome /

beneficiaries

Partnerships that address availability

O3B /

Google

Satellite mobile backhaul provided by

low earth orbit satellites at 8000km to

IAPs

Business case

for network roll-

out

Lower roll-out costs for

IAPs

Better availability for


Ref: 2004569-383

CAP Overview Barrier

addressed

Outcome /

beneficiaries

end users

Google Project Loon, an innovative project to

provide LTE coverage using

atmospheric balloons in partnership with

IAPs who provide radio spectrum

Business case

for network roll-

out


IAPs


end users

Google Titan Aerospace, which uses solar-

powered aircraft to provide broadband

capacity to end users

Business case

for network roll-

out


IAPs


end users

Facebook Provision of satellite broadband access

coverage in sub-Saharan Africa in

partnership with Eutelsat

Business case

for network roll-

out


IAPs


end users

Facebook Project Aquila which also uses solar-

powered aircraft to provide broadband

capacity to end users

Business case

for network roll-

out


IAPs


end users

Partnerships that address take-up

Mozilla Low-cost smartphones using the open

source Firefox operating system, made

in partnership with local Asia-Pacific

manufacturers

Device cost and

availability

Lower costs for end

users

More customers for

IAPs and CAPs

Google Android One programme to develop low-

cost smartphones in partnership with

local Asia-Pacific manufacturers

Device cost and

availability

Lower costs for end

users

More customers for

IAPs and CAPs

Facebook Free Basics programme to provide

limited zero-rated Internet access in

partnership with IAPs in the region

Social factors,

including

awareness

Greater awareness

and lower costs for

end users

More take-up for IAPs

Partnerships that address usage

Mozilla Mozilla handset localisation programme,

that provides local language handsets

and content in Asia-Pacific countries

Relevant content

in local

languages

More relevant content

for users

Greater usage for

CAPs

Google Android carrier billing, which integrates

the Google Play store with IAPs’ billing

systems, allowing customers to pay for

apps and content on their mobile bill

Relevant content Easier access to

relevant content for

users

More usage revenue

for CAPs and IAPs

Facebook Partnership with Ericsson and Axiata to

improve network quality of service in

Indonesia

Quality of

service

Better experience for

users

More usage revenue

for CAPs and IAPs


Ref: 2004569-383

0.3.2 Proactive policy makers are implementing innovative policy initiatives to lower barriers,

often in partnership with IAPs or CAPs

Policy makers are also working to address barriers, often in partnership with IAPs or CAPs. Here, we

present an overview of a number of policy initiatives that are lowering barriers in the Asia-Pacific

region (shown in Figure 0.11), setting out a description of the barrier addressed and the outcome.

Figure 0.11: Overview of example policy initiatives [Source: Analysys Mason, 2015]

Country Initiative Barrier addressed Outcome / beneficiaries

Initiatives that address availability

Japan Public funding for

broadband roll-out

Business case for

network roll-out

Increased take-up of fibre,

benefiting IAPs, CAPs and

users

South Korea Public funding for

broadband roll-out

Business case for

network roll-out



users

Singapore Development of national

broadband network

Business case for

network roll-out



users

Initiatives that address take-up

South Korea ICT education

programme

Social factors,

including awareness

of the benefits of the

Internet

Improved awareness for

~16 million people,

potentially improving take-up

and usage

Malaysia 1Malaysia netbook

distribution programme

Affordability of

connected devices

PC penetration increased to

~65%, lowering a key barrier

for users to take up

broadband

Initiatives that address usage

Japan Investments in e-

government

Relevant accessible

content

More efficient and relevant

content and applications for

users

Malaysia Investments in

e-government, e-education

and e-health

Relevant accessible

content



users

0.3.3 Policy makers have an essential role to play in ensuring that barriers to innovation and

investment are as low as possible

In addition to public interventions discussed above, policy makers play a crucial role in lowering

Internet barriers by creating an attractive regulatory environment. Here we discuss how good

public policy practices can help to achieve this. We draw lessons from the outcomes achieved by


Ref: 2004569-383

countries in the ‘connected’, ‘connecting’ and ‘unconnected’ clusters, and from our discussion of

key barriers in Section 3.3.17

We have identified four examples of good public policy practice.

Sector

liberalisation and

the encouragement

of sustainable

development

This includes a licensing regime that does not unnecessarily limit the

number of new entrants and has few restriction on technology or services

deployed. Scarce resources such as radio spectrum should be managed

transparently, so that as much spectrum as possible is available to market

participants including non-traditional users, to encourage innovation and

new investment.

Effective

competition

regulation that

stimulates

investment

In many of the countries of interest in this study, consumers are enjoying a

great amount of choice and operators are competing vigorously for

customers. Competition regulation should strike a balance between

encouraging competition, investment and innovation. Regulatory

forbearance should be applied where competition issues are unclear or

transitory. This is often the case for online content and platforms, which are

enablers of content consumption and innovation, and for new, risky

investment by established operators.

Attractive business

conditions

Business and investment conditions should be conducive to investment.

This includes low barriers to foreign direct investment, and a light

administrative burden to starting and running a business. The tax burden on

telecoms services and devices should not be excessive. Censorship and

surveillance of Internet users and content providers can discourage

investment.

Policies should be in place that lower barriers to the construction of fibre

networks. This includes policies that encourage passive infrastructure such

as ducts and poles to be shared, streamline construction permissions, and

ensure newly installed ducts are large enough to be shared.

Carefully balanced

data protection and

localisation

regulations

Data protection and localisation regulation should be carefully traded off

against the encouragement of innovation. Requirements that strictly limit

the purpose for which data can be used may inhibit innovation by

preventing CAPs using hidden patterns in databases to create new services.

Likewise limitations on the transfer of data between firms or countries may

inhibit innovation by preventing firms from combining their capabilities to

develop new services.

17 This builds on work we have conducted for regulators and operators in the region (including the GSMA), and for

NGOs and industry groups such as the Internet Society in Africa and Latin America.


Ref: 2004569-383

Forced data localisation policies, such as those that require certain data to be

held in-country, raise a number of concerns. They can have negative

impacts on consumers and the Internet ecosystem, by raising costs and

restricting access to services. They are difficult to implement through

regulation, as they require substantial underlying infrastructure to be present

in the first place to work in practice. These policies can also deter

investment in other data-heavy industries, including banking and retail,

which are essential to the development cycle of emerging economies.

Finally, they are also not very good at preserving privacy and security; in

fact, local storage policies may be counter-productive if they increase the

vulnerability of the data to cyber-attacks and reduce the resilience of

services that rely on it.

A more-effective policy is to ensure that local market conditions are

attractive for investment by providing high-quality, low-cost infrastructure,

and following the good practices outlined here.


Ref: 2004569-383

1 About this report

This report explores how the broadband ecosystem has developed in Asia–Pacific over the last 15

years, and the opportunities that this has created for consumers and telecoms operators throughout

the region. It draws upon Analysys Mason’s extensive experience of working in Asia–Pacific, for

operators, policy makers, regulators and investors, as well as an original consumer survey

conducted in five countries, spanning over 4000 respondents.

Our analysis covers 21 countries: Australia, Bangladesh, Brunei Darussalam, Cambodia, Hong

Kong, Indonesia, Japan, Laos, Malaysia, Myanmar, Nepal, New Zealand, Pakistan, Papua New

Guinea, the Philippines, Singapore, South Korea, Sri Lanka, Taiwan, Thailand and Vietnam.18

These countries are highly varied, ranging from the most advanced broadband markets in the

world (Japan and South Korea) to the least advanced (e.g. Papua New Guinea, Bangladesh).

In Section 2 of this report, we show that the tremendous growth in broadband (28% annual growth

since 2000) has created opportunities for the telecoms operators that provide these broadband

services in the region. As convergence has moved from buzzword to reality, these Internet access

providers (IAPs) have kept investing in modernising, upgrading and expanding their networks,

generating returns that are now overwhelmingly driven by broadband services. They have also

begun to take advantage of new opportunities created by the increasing spread of advanced

broadband19

, by deploying more-advanced networks and innovating with services, including

“over-the-top” (OTT) online content and services.

In Section 3, we highlight how Internet content and application providers (CAPs), whose core

business is content, have helped to drive this growth and enable these new opportunities.

Fundamentally, consumers take up broadband services in order to access advanced online content.

This content exists because of the investments made by CAPs, both large and small. Likewise, the

investments that CAPs have made in networks and infrastructure have helped to improve the

experience of consumers when using the Internet. This also helps to drive broadband take-up.

Finally, in Section 4 we show how broadband take-up and usage throughout the region, including

in the less connected countries, could be accelerated through innovative partnerships, good

regulation and judicious public initiatives.

This report includes a number of annexes containing supplementary material:

18 India and China are not covered in this report, as their scale and diversity means they warrant separate reviews of

their own.

19 We define advanced mobile broadband as services delivered over either a 3G or 4G network. We define advanced

fixed broadband as a service to a fixed location offering download speeds of 4Mbit/s or greater, based on the approximate connection speed required to watch a single standard-definition video stream. Other policy bodies have their own definitions. For example, the ITU defines broadband as a 2Mbit/s connection, based on the speed offered by primary-rate ISDN, while the FCC recently changed its definition of broadband from 4Mbit/s to 25Mbit/s.


Ref: 2004569-383

Annex A sets out the methodology of primary research we have conducted

Annex B discusses the current state of the broadband ecosystem in the region

Annex C describes the key barriers to broadband development in the region.



2 The rise of broadband Internet has created opportunities for

Asia–Pacific telecoms operators to thrive

The rapid growth in take-up and usage of advanced broadband in Asia–Pacific has created

opportunities for players across the value chain. Despite a shift in demand away from legacy voice

services towards broadband Internet services, IAPs in particular have seen opportunities to make

sustainably profitable investments in advanced networks. This shift has also created opportunities

for IAPs to ‘move up’ the value chain, and make investments in new and innovative OTT services.

In this section, we explain how IAPs are seizing these opportunities, generating sustainable returns

and investing in future growth opportunities.

2.1 Broadband Asia-Pacific: a strong growth story, but wide variations remain

Internet markets in Asia–Pacific have seen strong growth over time, in terms of both take-up of

broadband services and the use of online content and applications, as shown in Figure 2.1 and

Figure 2.2.

Figure 2.1: Total 3G and 4G connections in Asia–

Pacific [Source: GSMA, 2015]

Figure 2.2: Total annual data usage (PB) in Asia–

Pacific [Source: Analysys Mason, 2015]

Although the region as a whole has seen strong growth, the level of broadband development varies

widely from one country to another. We have classified the 21 countries of interest into three

groups, or ‘clusters’, reflecting varying stages of development:

Connected countries. These countries are advanced, with near-ubiquitous coverage of 3G and

4G mobile and fixed broadband, high take-up, and high usage of advanced services. Countries

0

80

160

240

320

400

200

5

200

6

200

7

200

8

200

9

201

0

201

1

201

2

201

3

201

4

Mill

ion

CAGR = 28%

0

20

40

60

80

100

201

2

201

3

2014

Th

ou

sa

nd

CAGR = 28%


Ref: 2004569-383

in this cluster are Australia, Hong Kong, Japan, New Zealand, Singapore, South Korea and

Taiwan.

Connecting countries. These countries have high coverage, but moderate take-up of 3G and

4G mobile broadband, low coverage and take-up of fixed broadband, and increasing usage of

advanced Internet services. Countries in this cluster are Brunei Darussalam, Cambodia,

Indonesia, Malaysia, the Philippines, Sri Lanka, Thailand, and Vietnam.

Unconnected countries. These countries have yet to develop advanced broadband services.

They have very little coverage or take-up of either 3G mobile services or advanced fixed

broadband services. Countries in this cluster are Bangladesh, Laos, Myanmar, Nepal, Pakistan,

and Papua New Guinea.

Figure 2.3 below shows where these countries sit in the Networked Readiness Index produced by

the World Economic Forum20

.

Figure 2.3: Country clusters against Networked Readiness Index [Source: World Economic Forum, 2015]

For more detail regarding the broadband markets in these clusters, please refer to Annex B.

20 This index measured the level of development of the ICT sector, and includes metrics such as the regulatory

environment, quality of ICT infrastructure, usage by business and consumers, and social benefits of ICT.

0

1

2

3

4

5

6

7

Netw

ork

ed

Rea

din

ess In

de

x

Countries

‘Connected’ ‘Connecting’ ‘Unconnected’


Ref: 2004569-383

2.2 The current picture: broadband is replacing legacy services as the main revenue and

profit driver for IAPs

In this context of strong growth, the core business of IAPs has evolved from being voice- and

private data-centric, to being focused on selling broadband access services. This has sometimes

resulted from a shift in demand from existing narrowband services towards broadband, and indeed

from other ‘legacy’ services such as voice telephony and SMS, which consumers can in part

replace by over-the-top (OTT) services using a broadband connection.

Overall, IAPs in the region have seen their revenue grow as broadband has developed,21

but this

has not translated into higher profits always and everywhere, in part because of these shifts from

old to new services and in part because of increased competition in many markets. To provide

these new services, however, IAPs have continued to make investments and have managed to

maintained a relatively stable rate of returns on these investments, avoiding commoditisation and

price declines even as the industry matured. In addition, IAPs have begun to invest in innovative

OTT services, which have offered opportunities for growth beyond the traditional network

business. We explore those further in Section 2.3.

Asia–Pacific telco revenues have seen sustained growth since the introduction of broadband

services

The impact on IAPs of the strong growth in broadband is clear from the sustained growth in

overall telco revenues throughout the region since the introduction of advanced broadband, as

shown in Figure 2.4. This shows total industry revenue growth from fixed and mobile services

since the introduction of advanced broadband. It is clear that both ‘connected’ and ‘connecting’

countries have seen strong growth since the development of broadband began.

Note: ‘unconnected’ countries are excluded from this analysis, as these have yet to develop

advanced broadband.

21 In both real and nominal terms, i.e. growth has outstripped inflation


Ref: 2004569-383

Figure 2.4: Overall IAP

revenue in ‘connected’

and ‘connecting’

clusters, indexed to the

year of 3G launch,

using constant

exchange rates

[Source: GSMA,

company data, ITU,

TeleGeography,


In addition to an overall increase in revenue, IAPs have also seen a very marked increase in the

importance of data and Internet services within their revenue mix. In all countries, the proportion

of industry revenue from data and Internet services has increased significantly since the launch of

advanced broadband services. This is illustrated in Figure 2.5 below, starting from the year of 3G

launch. Connecting countries have rapidly caught up with ‘connected’ countries, demonstrating the

importance of the Internet in the telecoms sectors of many middle-income Asia–Pacific

economies.


of total service revenue

from data and Internet

services [Source:

GSMA, company data,

ITU, TeleGeography,


0

50

100

150

200

250

300

350

200

1

2002

200

3

200

4

200

5

2006

200

7

200

8

200

9

201

0

201

1

201

2

201

3

201

4

Re

ve

nu

e in

de

xe

d t

o 1

00


Connected 3G launch Connecting 3G launch

0%

10%

20%

30%

40%

50%

60%

70%

0 1 2 3 4 5 6 7 8 9 10 11 12 13

Pro

po

rtio

n o

f re

ve

nu

e fro

m d

ata




Ref: 2004569-383

Profit growth is now largely attributable to broadband

As revenue has grown and shifted from voice to data services (including Internet access), IAP

profitability22

has shifted from traditional services such as voice telephony to data services,

including broadband Internet access and usage. Over this period, EBITDA has been constrained,

but relatively stable in all countries in the region (reflecting many factors such as subsidies for

high-end handsets). Figure 2.6 below shows growth in total industry EBITDA, for both fixed and

mobile services, since the first introduction of advanced broadband networks.

Figure 2.6: Total market

EBITDA indexed to the

year of launch of 3G

[Source: GSMA,

company data, ITU,

TeleGeography,


With modern networks, IAPs are able to provide all their services over the same infrastructure, and

so the shift in revenue towards data and Internet services implies that these are now a significant

(and in some cases the dominant) source of IAP profits. Key to this are developments in 4G

mobile access networks, which are fundamentally broadband data networks that allow spectrum

and access network resources to be shared seamlessly between services. Likewise, the

development of converged core networks allows all voice and data traffic to be carried seamlessly

over an IP network.

This technical evolution has increased economies of scope by removing the need to have dedicated

equipment for both voice and data services, and has also increased economies of scale by

improving the utilisation of the shared equipment. As a result, the profits made by IAPs are now

mainly attributable to data services, a pattern that is set to continue for years to come. Figure 2.7

22 We use earnings before interest, taxes, depreciation and amortisation, or EBITDA, as a proxy here.

0

20

40

60

80

100

120

140

160

180

0 1 2 3 4 5 6 7 8 9 10 11 12 13

EB

ITD

A in

de

xe

d to

10

0




Ref: 2004569-383

below illustrates very simply what profits from broadband could look like on the basis of their

share of revenue.23

Figure 2.7: Illustrative

EBITDA, overall and

from data services,

constant exchange

rates [Source: GSMA,

company data, ITU,

TeleGeography,


Broadband investment opportunities and the returns on these investments have remained strong

To achieve this growth in data revenue and profit, IAPs have made significant investments in their

networks. The majority of these investments have been in the access networks that connect to end

users, as well as in core networks that can seamlessly handle all types of service. Mobile operators

have made investments in rolling out advanced 4G networks. Fixed operators have invested in

either upgrades to network equipment (in the case of cable operators and traditional telecoms

operators investing in DSL networks) or the installation of new fibre links to end customers.

The magnitude of these investments is significant at between USD17 billion and USD19 billion

per annum in the region over the past five years, as shown in Figure 2.8 below for the ‘connected’

and ‘connecting’ clusters.

23 This is highly simplified, but in the context of converged, multi-services networks, it does provide a reasonable first-

order estimate of profit allocation

0

20

40

60

80

100

120

140

160

180

0 1 2 3 4 5 6 7 8 9 10 11 12 13

EB

ITD

A in

de

xe

d to

10

0


Connected ConnectingConnected: from data Connecting: from data


Ref: 2004569-383

Figure 2.8: Total annual

capital expenditure in

‘connected’ and

‘connecting’ clusters, at

constant exchange

rates [Source: Analysys



ITU, 2015]

These investments have continued to produce healthy returns, as profits from Internet and data

services have held up well throughout the region. Overall industry returns, measured by the return

on capital employed (ROCE) ratio,24

are illustrated in Figure 2.9 below. These appear healthy

overall for mature markets, comparable to estimates of the cost of capital for telecoms operators

(represented by the green band in Figure 2.9).25

Countries where network upgrades are recent or

currently underway have a lower level of ROCE partly because equipment costs tend to be larger

compared to their revenues.

24 ROCE is defined as earnings before interest and tax divided by total capital employed.

25 Compiled based on data by Aswath Damodaran, NYU Stern, 2015; the range shown is for telecoms services and

wireless providers, and comprises Global, Emerging markets and Japan values.

0

5

10

15

20

25

2010 2011 2012 2013 2014

US

D b

illio

n



Ref: 2004569-383

Figure 2.9: ROCE for

IAPs in ‘connected’ and

‘connecting’ clusters

[Source: Analysys



ITU, Aswath

Damodaran (NYU

Stern), 2015]

* The ROCE for Thailand is for mobile only.

2.3 Investing in the future: IAPs in Asia–Pacific are increasingly leveraging their

customer relationships to drive adoption of innovative OTT services

In addition to the returns from investment in broadband networks, IAPs have begun to take

advantage of the opportunities to invest in new and innovative OTT services. Investments have

been made by IAPs in both developing and developed countries, in a range of verticals, including

video and entertainment, financial services, advertising and broader applications platforms. Whilst

these initiatives are in their early stages, there is evidence of IAPs in the region achieving growth

in these areas.

OTT investments have been made by IAPs in both developing and developed countries, in a range

of verticals, including video and entertainment, financial services, advertising and broader

applications platforms. Examples from developing countries include:

Singtel’s investments in OTT services, including the Amobee mobile advertising platform and

the Hooq OTT video service in Thailand, India, and the Philippines

PLDT’s investments in payment and remittance services, and in the Smart Pinoy app, which

provides content and services to Filipinos overseas

Grameenphone’s investment in WowBox, a zero-rated26

mobile content app aimed at prepaid

users, funded by advertising.

26 Content that is zero-rated does not attract usage charges from the IAP or deplete a customer’s monthly usage

quota.

0%

5%

10%

15%

20%

25%

Au

str

alia

Hon

g K

on

g

Ne

w Z

ea

land

Sin

ga

pore

Ta

iwa

n

Ma

laysia

Jap

an

Sri

La

nka

Th

aila

nd

*

So

uth

Ko

rea

Vie

tna

m

Indo

nesia

Ph

ilipp

ines

RO

CE



Ref: 2004569-383

Examples from developed countries include:

NTT DoCoMo’s investments in the ‘dmarket’ mobile application portal

KDDI’s Smart Pass service, which allows access to the top-500 Android apps for a monthly

fee

Singtel’s partnership with Fetch TV, a subscription-based OTT television service, via its

Australian subsidiary, Optus, and its TV GO service in Singapore.

We explore these in more detail below, looking at how the IAPs are leveraging their existing

capabilities, such as their networks and relationship with a large customer base, and then

discussing the role played by content and application providers (CAPs).

In developing countries, IAPs leverage their customer relationships and networks to reach

customers underserved by global CAPs

In developing countries, IAPs are helping to bring online services and content that are highly

tailored to specific audiences, in a way that is not easily doable for global CAPs. For example,

PLDT is targeting overseas Filipino workers, and Grameenphone targets price sensitive prepaid

customers with advertising funded content. In doing so, IAPs take advantage of their existing

relationships with the vast majority of end-users in their markets. They are doing this in a number

of ways:

They leverage their customer relationships and marketing knowledge to provide seamless

services targeted at specific national groups. For example, PLDT’s Smart Money and Smart

Pinoy services target overseas Filipino workers, a large group across Asia whose needs are not

well served by either global CAPs or non-Philippine IAPs.

They leverage their customer databases and billing relationships to provide targeted

advertising services. Grameenphone’s WowBox application and Singtel’s Amobee mobile

advertising platform are both examples of this.

It should be noted that CAPs are involved as partners in a number of these investments, where they

provide capabilities or assets that IAPs do not have. For example, Singtel’s Hooq service relies on

content produced by major studios (Sony Pictures and Warner Bros), and IAP applications are

promoted and distributed through online and mobile app stores.

An overview of selected OTT opportunities in developing countries is provided in Figure 2.10.

This summarises how each opportunity leverages the existing capabilities of IAPs, and where

relevant outlines the involvement of CAPs.


Ref: 2004569-383

Figure 2.10: Selected OTT opportunities for IAPs in developing countries [Source: Company data, Analysys

Mason, 2015]

Overview of opportunity Leverage of existing

IAP capabilities CAP involvement

Singtel

Amobee is a mobile advertising

platform which provides real-

time mobile ad serving, and

detailed reporting and ad

response analysis to

advertisers

Leverages data from Singtel’s

large existing mobile customer

base, allowing better targeting

of advertising

N/A

The Hooq OTT video app

allows video content to be

streamed to any device on a

Singtel or partner network. It

has been launched in Thailand,

India and the Philippines

Leverages existing customer

base for marketing and billing

relationship for ease of

payments.

Content partnership with

Warner Bros and Sony

Pictures. Relies on app store

for distribution

PLDT

Smart Money is an electronic

wallet which allows bill

payment, money transfer to

other Smart mobile

subscribers, and remittances

from overseas using a Smart

mobile phone

Takes advantage of Smart’s

large mobile customer base in

the Philippines (51 million), and

the large numbers of Filipino

overseas workers who send

remittances back home

Smart Money is accepted in

Apple’s App Store and on

iTunes

Smart Pinoy is an e-commerce

platform targeted at overseas

Filipino workers. It allows

customers to send gifts to

friends and family in the

Philippines, as well as

providing other services such

as free calls to consulates

Takes advantage of Smart’s

brand strength among

overseas Filipinos, as well as

distribution networks in the

Philippines

Smart Pinoy is distributed

through CAP app stores

Grameenphone

The WowBox app provides

mobile content to prepaid

users that is funded by

advertising. Use of WowBox

does not deplete users’

prepaid balance

Allows Grameenphone’s large,

price-sensitive prepaid

customer base to access

Internet content free of charge

N/A

In developed markets, IAPs are distributing advanced content-rich OTT services, often together

with CAPs but sometimes on their own platforms

Consumers in developed markets already have access to a wealth of advanced content and

applications on their fixed and mobile devices. Because of this, IAPs are using their customer

relationships, sales channels, billing relationships and networks to act as distributors of advanced

content. They do this either by partnering with CAPs (e.g. KDDI’s Smart Pass, or Optus with

Fetch TV) or over their own platforms (e.g. NTT DoCoMo’s dmarket, Singtel’s TV GO service).


Ref: 2004569-383

An overview of selected OTT opportunities in developed countries is shown in Figure 2.11.

Figure 2.11: Selected OTT opportunities for IAPs in developed countries [Source: Company data, Analysys

Mason, 2015]

Overview of opportunity Leverage of existing

capabilities CAP involvement

NTT DoCoMo

The dmarket application

platform comes pre-installed on

NTT DoCoMo smartphones. It

supports the download of

content (e.g. anime, TV) and

offers a variety of applications

Leverages NTT DoCoMo’s

large customer base and wide

range of Japan-specific content

and applications

dmarket competes with CAP

application platforms

KDDI

Smart Pass comes pre-

installed on KDDI Android

devices. It provides unlimited

access to the top-500 Android

apps for a monthly subscription

fee

Uses KDDI’s billing relationship

with its customers to make it

simpler for users to consume

content and applications, thus

helping to drive data usage

Service has been developed in

partnership with Google /

Android

Singtel

Singtel’s Australian subsidiary,

Optus, offers Fetch TV on its

fixed and mobile networks.

Fetch TV is a subscription-

based OTT TV service that

provides TV channels in

English and several Asian

languages, as well as Netflix. It

is viewed via a set-top-box or

mobile app

Uses Optus’s fixed and mobile

broadband networks, and

leverages its billing

relationships with customers,

and marketing and sales

channels

Fetch TV is offered by a CAP

that is backed by Malaysia’s

Astro (a major pay-TV

operator)

Singtel TV GO is an OTT

mobile video app. It is offered

to Singtel’s existing pay-TV

subscribers free of charge

Uses Singtel’s customer

relationships, content deals

and network to offer pay-TV

customers a differentiated

service

Singtel has deals with a

number of content owners to

distribute pay-TV content to

fixed and mobile devices

IAPs can offer compelling OTT services where they can leverage their existing capabilities and assets

IAPs have a number of capabilities and assets that they can use to carve out a space in the rapidly

growing market for OTT content and applications. In particular, they have a deep existing

relationship with their existing customers, including sales channels, marketing knowledge and

billing arrangements. In both developed and developing countries, there are examples of IAPs

leveraging these capabilities in their OTT investments.

Despite the early-stage nature of most of the investments, the fact that IAPs are making them

indicates the opportunities that are being created by the increased availability of advanced

broadband. As the Internet reaches more people in the region, and the number of users continues to

grow, these opportunities are likely to increase.

Broadband in Asia–Pacific: How investment, partnerships and policy are driving a global success story | A–32


3 CAPs are helping to drive growth and opportunities for

IAPs and other participants in the Internet value chain

CAPs create the content and services available on the Internet, and have been key drivers of

growth and enablers of opportunities in the broadband world. This has benefited all players in the

value chain by driving take-up and usage of broadband services.

Consumers take up or upgrade their broadband services in order to access advanced online content.

When they do not take up a broadband service, the most significant factor is either a lack of

content relevant to them, or a lack of awareness about this content. The content which exists does

so because of the investments made by CAPs. In addition, the investments made by CAPs in

networks and infrastructure have helped to improve users’ experience of the Internet. This also

helps to drive take-up of broadband services.

In ‘unconnected’ countries, there are further barriers to Internet development which have an

arguably greater impact than that seen in the other clusters. These include affordability, a lack of

local-language content, and PC/smartphone penetration, among others. CAPs large and small are

already investing in trying to alleviate some of these barriers, and exploring partnerships with IAPs

to address them (something we discuss in more depth in Section 4).

We discuss each of these issues below.

3.1 Consumers take up broadband services explicitly to access content and services

created by CAPs

For adopters of broadband, online content and applications are the main drivers behind usage,

the decision to upgrade their access and the value they get from the Internet

The main reason why consumers sign up for broadband access from IAPs is to use the content and

services created by CAPs. For example, research conducted in 2013 for the Computer and

Communications Industry Association (CCIA) found that European consumers who use advanced

services are more likely to upgrade their connection: 40% of those making video calls and 49% of

those uploading video upgraded in the year preceding the study.27

New research we commissioned as part of this study, focusing on Asia-Pacific, shows similar

conclusions28

. Fixed and mobile Internet subscribers in the five countries surveyed make very

27 See http://www.ccianet.org/2013/09/new-research-shows-high-bandwidth-online-services-create-demand-for-faster-

more-expensive-broadband/.

28 Analysys Mason has conducted primary research among Internet adopters in Australia, Indonesia, Singapore,

Sri Lanka and Thailand, and among non-adopters in Indonesia, Sri Lanka and Thailand. For more details of the methodology and findings of this research, please refer to Annex A.


Ref: 2004569-383

frequent use of key Internet services offered by CAPs: 88% of fixed users (Figure 3.1) and 82% of

mobile users (Figure 3.2) access social networks either every day or several times a week, and

80% of fixed and 66% of mobile users access video content daily or several times a week.


connection for the following services? (n=2202)

[Source: Analysys Mason, 2015]


Internet connection for the following services?

(n=4100) [Source: Analysys Mason, 2015]

As people use the Internet, their appetite for richer content increases, which drives their interest in

more-advanced broadband access services. We found that between 61% and 88% of fixed

customers and between 52% and 87% of mobile customers would consider upgrading their

services, for reasons of either price, quality or both (illustrated in Figure 3.3 and Figure 3.4.)

0% 50% 100%

Listening to music

Playing games

Browsing

Watching videos

Communications

Social networking



Very rarely Never

0% 50% 100%

Listening to music

Playing games

Watching videos

Communications

Browsing

Social networking



Very rarely Never


Ref: 2004569-383

Figure 3.3: What would make you consider

upgrading your current fixed Internet service?

(Australia n=529, Indonesia n=375, Singapore

n=677, Sri Lanka n=194, Thailand n=427) [Source:


Figure 3.4: What would make you consider

upgrading your current mobile Internet service?

(Australia n=1000, Indonesia n=800, Singapore

n=1000, Sri Lanka n=500, Thailand n=800) [Source:


This strong potential intention to upgrade suggests that there may be a latent demand for better

services. Indeed, between 29% and 49% of fixed users and between 25% and 50% of mobile users

responded that they would upgrade in order to receive a better user experience.

In addition, the majority of current Internet users who stated that they would upgrade their fixed

and mobile services (‘upgraders’) are significantly more likely to use advanced services such as

video:

69% of mobile upgraders (Figure 3.5) use video several times a week or more, compared to

57% of non-upgraders (Figure 3.6).

82% of fixed upgraders (Figure 3.7) use video regularly, compared to 76% of non-upgraders

(Figure 3.8).

0% 50% 100%

Australia

Indonesia

Singapore

Sri Lanka

Thailand

More affordable

Better user experience

Both of the above

Don't want to upgrade

Already on the best plan

I don’t know

Other

0% 50% 100%

Australia

Indonesia

Singapore

Sri Lanka

Thailand

More affordable

Better user experience

Both of the above

Don't want to upgrade

Already on the best plan

I don’t know

Other


Ref: 2004569-383



(asked of mobile upgraders,29

n=2894) [Source:




(asked of mobile non-upgraders, n=1206) [Source:



connection for the following services? (asked of fixed

upgraders,30

n=1696) [Source: Analysys Mason,

2015]


connection for the following services? (asked of fixed

non-upgraders, n=506) [Source: Analysys Mason,

2015]

29 Mobile upgraders are defined as mobile users who gave a response other than “I don’t want to upgrade”, “I am

already receiving the best connectivity available at my location”, “Other”, or “Don’t know” to the question “What would make you consider upgrading the current mobile Internet service?”.

30 Fixed upgraders are define as mobile users who gave a response other than “I don’t want to upgrade”, “I am already

receiving the best connectivity available at my location”, “Other”, or “Don’t know” to the question “What would make you consider upgrading the current fixed Internet service?”.

0% 50% 100%

Listening to music

Watching videos

Communications

Playing games

Browsing

Social networking



Very rarely Never

0% 50% 100%

Watching videos

Listening to music

Communications

Playing games

Social networking

Browsing



Very rarely Never

0% 50% 100%

Listening to music

Playing games

Watching videos

Communications

Social networking

Browsing



Very rarely Never

0% 50% 100%

Listening to music

Playing games

Watching videos

Communications

Social networking

Browsing



Very rarely Never


Ref: 2004569-383

This usage of online content and services delivers value to users. Users receive value (known as

‘consumer surplus’) from a service if the maximum amount they would be prepared to pay (their

‘willingness to pay’) is greater than the market price for the service.

In our research we examined users’ willingness to pay for Internet services, relative to what they

currently pay (Figure 3.9). For fixed services, we found that upgraders have a higher willingness to

pay than non-upgraders, although this relationship is weaker for mobile users. This suggests that,

at least for fixed users, the drive to upgrade, itself linked to the use of advanced services, is linked

to a higher willingness to pay.

In addition, it appears that the willingness to pay of upgraders in ‘connecting’ countries (here

Thailand, Sri Lanka, and Indonesia) has a greater premium over non-upgraders when compared to

‘connected’ countries (Australian and Singapore). This means that if these countries could drive

upgrades, potentially by lowering barriers to network investments that could improve quality of

service, there could be significant value created by unlocking this willingness to pay. We discuss

initiatives and policies that can achieve this in Section 4.

Figure 3.9: Premium

that fixed users are

willing to pay, above the

price they currently pay

(Australia n=404,

Indonesia n=297,

Singapore n=520, Sri

Lanka n=137, Thailand

n=376) [Source:

Analysys Mason, 2015]]

If we look at the average price paid for narrowband mobile services (before the introduction of

broadband 3G and 4G networks), and see how it has changed as customers have shifted to

broadband, and also look at how customer numbers have grown, we can gain an understanding of

whether overall consumer surplus has increased. Looking at Figure 3.10 and Figure 3.11 below, it

is evident that prices have decreased substantially as subscriber numbers have grown. As

willingness-to-pay for broadband is significantly higher than for narrowband, thanks to a richer

user experience, ‘consumer surplus’, a key measure of economic welfare, has increased

significantly.

0% 40% 80% 120%

Australia

Indonesia

Singapore

Sri Lanka

Thailand

Fixed Non-upgraders Fixed Upgraders


Ref: 2004569-383

Figure 3.10: Weighted average ARPU, pre- and

post-broadband, at constant exchange rates,

countries in all clusters [Source: GSMA, 2015]

Figure 3.11: Total mobile Internet subscribers, pre-

and post-broadband, countries in all clusters

[Source: Analysys Mason, 2015]

Creating and raising awareness of content and applications that are relevant and accessible to

non-adopters is the most important factor in improving Internet take-up

As part of our primary research programme, we spoke to ‘non-adopters’31

in Indonesia, Sri Lanka

and Thailand to discuss their motivations and understand what they perceive as the main reasons

why they do not adopt or use the Internet (Figure 3.12).

The most important reason cited was “awareness”,32

mentioned by 75% of respondents. This is a

measure of whether consumers are aware of the benefits of consuming Internet content.33

Service

availability, which is driven by access to a fixed or mobile broadband network, was substantially

less important. Literacy and the lack of a connected device were less important than awareness, but

did appear to be a barrier for a significant minority of people. The availability of relevant content

was also a barrier for a significant minority.

These results are also borne out by other research. A 2014 in Singapore by the national regulatory

authority, IDA34

, found that, of non-adopters, 55% said they did not need an Internet connection,

and 30% said they did not know how to use the Internet.

31 Defined as people who have neither a personal mobile Internet service nor a fixed Internet service at their place of

residence.

32 Defined as respondents who provided one of the following answers: “I don’t trust it / don’t think it’s safe”, “I don’t

need it” or “I don’t know what the Internet is or what it does”.

33 It is possible that well informed people may not want the Internet. Evidence from connected countries, however,

which have near-ubiquitous take-up, shows that this group is small.

34 IDA Consumer Awareness and Satisfaction Survey, 2014

0

10

20

30

40

50

60

Narrowband Broadband

US

D p

er

mo

nth

0

50

100

150

200

250

300

350

400

Narrowband Broadband

Mill

ion

s


Ref: 2004569-383

Figure 3.12: Reasons cited for non-adoption or non-usage of Internet services as stated by the interviewees

in all countries [Source: Analysys Mason, 2015]

Notably, the majority of non-adopters have never even used the Internet before, even at an Internet

café, at the home of a friend or family member, or at work or an educational institution

(Figure 3.13). This lack of experience of the Internet is also likely to be associated with a lack of

awareness of its benefits.

Figure 3.13: Responses

from non-adopters on

their use of an Internet

service (Indonesia

n=400, Sri Lanka

n=402, Thailand n=415)

[Source: Analysys

Mason, 2015]

Although most people would try the Internet if it were free (Figure 3.14), a surprisingly large

proportion of people who responded that they do not need the Internet said they would not be

0% 10% 20% 30% 40% 50% 60% 70% 80%

Other

It's too expensive

Someone close to me can use it for me when Ineed something

Availability

I don't own a connected device

Literacy

Relevance of content

Awareness

66%

82%76%

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Indonesia Sri Lanka Thailand

Yes, I access theInternet every day

Yes, I access theInternet up to severaltimes a week

Yes, I use it severaltimes a year

Yes, a few times

Yes, once

No, never


Ref: 2004569-383

prepared to try it even if it were offered free of charge. This reinforces the need for better

awareness and relevant content.

Figure 3.14:

Respondents interested

in trying a free Internet

service, among those

who answered “I don’t

need it” as their reason

for not adopting the

Internet (Indonesia

fixed n=165, mobile

n=130; Sri Lanka fixed

n=232, mobile n=239;

Thailand fixed n=262,

mobile n=242) [Source:


As a result, although cost is not necessarily seen as a barrier, even amongst very low-income

respondents, it is likely to be a secondary barrier to adoption in many cases once relevance can be

addressed and demand materialises. This is also visible from the gap between how much adopters

and non-adopters are willing to pay for an Internet service (Figure 3.15 and Figure 3.16). This

confirms that non-adopters are substantially more price sensitive, and/or see less value in Internet

services, than adopters.

0%

10%

20%

30%

40%

50%

60%

70%

80%


Fixed Mobile


Ref: 2004569-383

Figure 3.15: and Figure 3.16 : Proportion of household expenditure spent in Internet access services by

adopters and median of the maximum proportion of household expenditure that an Internet service could cost

for fixed non-adopters to take up the service [Source: Analysys Mason, 2015]

Fixed broadband (Indonesia, non-adopters n=41,

adopters n=338; Sri Lanka, non-adopters n=73,

adopters n=169; Thailand, non-adopters n=21,

adopters n=380)

Mobile broadband (Indonesia, non-adopters n=29,

adopters n=677; Sri Lanka, non-adopters n=39,

adopters n=386; Thailand, non-adopters n=25,

adopters n=660)

In summary, awareness of the benefits of consuming Internet content is a key factor, both for

people who do not take up an Internet service, and for people who have never used it. Either

interesting and relevant content is not available, or people are unaware of the benefits of the

content that is available. Price may be a secondary consideration once people have been persuaded

to try the Internet, but network availability is not seen as an issue, reflecting the fairly high

coverage of mobile broadband services in the countries surveyed.

In this context, future growth in the adoption of Internet access services by current ‘non-adopters’

could be driven by awareness-raising initiatives and investment in content that is relevant and

accessible to these new users. CAPs, potentially including some IAPs, will play a central role in

this evolution.

3.2 Quality of service drives take-up and usage, and CAPs are investing alongside IAPs

in network infrastructure to improve the user experience of the Internet

Quality of service is linked to service upgrades for both fixed and mobile users

The quality of service received by an end user is a key aspect of their enjoyment of the Internet,

and it also helps to drive users to upgrade their services. Indeed, users are more likely to upgrade

their service if they have had a good experience of consuming advanced content. In our research,

7%7%

3%

5%

3%

2%

0%

1%

2%

3%

4%

5%

6%

7%

8%


Adopters Non adopters

4%

4%

3%

3%

2%

1%

0%

1%

2%

3%

4%

5%


Adopters Non adopters


Ref: 2004569-383

we asked respondents about the quality of their experience when watching video content. Both

fixed and mobile upgraders were more likely than non-upgraders to be able to watch video with no

degradation in quality at least most of the time. This is illustrated in Figure 3.17 for fixed and

Figure 3.18 for mobile.

Figure 3.17: Can you watch video on your fixed

connection without degradation, asked of fixed

upgraders and non-upgraders (Upgraders n=1696,

Non-upgraders n=326) [Source: Analysys Mason,

2015]

Figure 3.18: Can you watch video on your fixed

connection without degradation, asked of mobile

upgraders and non-upgraders (Upgraders n=2894,

non-upgraders n=787) [Source: Analysys Mason,

2015]

CAPs are increasingly improving the quality of user experience, which stimulates adoption, usage

and upgrades of Internet access services

As well as creating content and applications, CAPs increasingly make investments in their own

networks and infrastructure. This is in addition to the substantial investments that IAPs are making

in upgrading their core and access networks.

CAPs are investing at all levels of their networks. Figure 3.19 illustrates the activities of CAPs

involved in the Internet value chain, and their corresponding investments.

83%

76%

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Upgraders Non upgraders

69%

56%

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Upgraders Non upgraders


Ref: 2004569-383

Figure 3.19: Illustrative Internet value chain [Source: Analysys Mason, 2015]

A large part of the investment made by CAPs is in hosting infrastructure, including data centres,

servers and storage devices. This infrastructure hosts and processes the content and applications

developed by CAPs, and houses nodes where content and applications are distributed and cached.

CAPs are increasingly making this investment in Asia–Pacific. For example, Google has built data

centres in Singapore and Taiwan, and Amazon Web Services (AWS), the cloud-computing arm of

Amazon, has data centres in Singapore, Sydney, Tokyo and Beijing. Meanwhile, Facebook is

reported to be investigating Asia–Pacific data-centre sites to add to its assets in the USA and

Europe.

CAPs also invest in transport networks. These are long-distance submarine or terrestrial fibre-

optic cables that carry Internet traffic around the world. Whilst IAPs and backbone operators are

large investors in these, CAPs are now also major contributors. Indeed, CAPs are investing in

major new cables in Asia–Pacific, including APX-East (between Australia and the USA),

FASTER (between Japan and the USA) and BBG (between South-East Asia, the Middle East and

Europe).

Finally, CAPs invest in delivery infrastructure, which carries traffic from long-distance transport

networks to IAP networks. This includes the points of interconnect where networks exchange

traffic (either at Internet exchanges or private points of interconnect). It also includes content

delivery networks (CDNs) that help to optimise traffic delivery, for example by shifting content to

hosting infrastructure that is closer to end users.

Some of the key infrastructure that CAPs contribute to in the region is shown in Figure 3.20. This

shows key data centre locations and the routes of the three submarine cables discussed above.

Ste

ps

Content and

apps

Store,

process,

servers

(hosting)

Transport

towards

customer

Delivery to

the IAP IAP to end

consumer

Ac

tivit

ies

Provision of

content and

applications

Provision of

computing

and storage

capacity

Transport of

content from

the CAPs to

the end users

Transfer and

optimisation

of traffic flow

between

networks

Provision of

last-mile

access to the

end user

Inve

stm

en

ts

Software

development

and content

production

Data-centre

buildings and

related

equipment

Submarine

and terrestrial

cables

Internet

exchanges

and related

facilities

Last-mile

fixed and

mobile

networks

CAP investments


Ref: 2004569-383

Figure 3.20: Map with

locations of key

infrastructure [Source:


CAP investments have an impact on IAPs in two ways:

The investments are partly substitutional: these investments that CAPs make replace

investments that would otherwise need to be made by IAPs. These investments provide

network capacity that is required for the successful operation of the Internet. Their main

impact on IAPs is to save them money, by removing the need for IAPs to make the

investments themselves.

The investments are partly incremental: these improve the quality of service that end users

experience when consuming content and applications. CAPs make these investments to attract

more users to their services, but the investments also benefit various other parties:

– Consumers benefit from the ability to consume content and applications at a high quality

of service, which increases consumer willingness to pay for Internet services, and thus

increases overall demand

– IAPs benefit from this growth in demand for their services

– CAPs also benefit, as they attract customers because their content and applications are

available at a high quality of service.

In a 2014 paper,35

Analysys Mason estimated the magnitude of the global investments by CAPs in

networks and infrastructure (made up of hosting, transport and delivery infrastructure). We

estimated global CAP investments to be USD33 billion per annum, of which USD7.7 billion was

35 See http://www.analysysmason.com/Research/Content/Reports/Content-application-provider-Internet-infrastructure-

Sept2014/.

UAE

Oman

<15

USA

APX-East

BBG

FASTER

Number of data centres

>30

Submarine cable systems:

15-30


Ref: 2004569-383

in Asia–Pacific. Investments in submarine and terrestrial cable infrastructure were estimated to

average USD2.7 billion per annum, of which USD700 million was in Asia–Pacific.

3.3 Challenges remain in bringing the benefits of broadband to more people and to


We have established in Section 3.1 that a lack of awareness of the benefits of the Internet is one of

the foremost drivers of non-adoption and non-usage. This creates risks associated with potential

demand, which can deter investment in infrastructure, especially in ‘unconnected’ countries. IAPs

may find it too risky to invest in new network coverage, given apparent demand for Internet

services is low. Here we explore these other barriers, and the impact they have had on the


There are a wide range of barriers that can have an impact on broadband development outcomes

In addition to the barriers discussed in Section 3.2, which apply largely in ‘connecting’ countries,

we have identified a wider range that can influence all aspects of broadband development. We can

define drivers and barriers to development with reference to the outcome that they influence:

The availability of broadband services determines whether end users are able to use these

services at all. It includes the coverage of advanced fixed and mobile access networks, as well

as the backbone, hosting and other infrastructure that must exist for content and applications to

be consumed at a high quality of service

The take-up of services reflects the extent to which consumers decide to become Internet

users and sign up for services from fixed and mobile operators

The usage of services refers to how much consumers actually use their Internet connections to

access content and applications.

For each of these outcomes, we have identified a set of potential drivers (if they are present) or

barriers (if they are lacking), on the basis of how IAPs address their respective markets, how

governments and regulators define and implement broadband policy, and how users decide to

subscribe to and use broadband services.

We set out an overview of the key barriers and drivers we have identified for each outcome in

Figure 3.21.

Figure 3.21: Drivers and barriers related to availability, take-up and usage [Source: Analysys Mason, 2015]

Driver /

barrier

Description

Drivers and barriers related to availability

Business

case

Is there an economic business case for rolling out broadband coverage somewhere new?

Is there a policy case for it and the money to fund it? Roll-out requires fixed and mobile

access networks, long-distance domestic and international backbone networks, and


Ref: 2004569-383

Driver /

barrier

Description

hosting and interconnection infrastructure.

Two key factors play a role here: the cost of inputs such as land, equipment, spectrum

and licences; and the risk associated with demand from consumers.

Availability

of inputs

Are the key inputs that required available at all? Scarcity of inputs affects three areas:

Radio spectrum and land are important for mobile networks (handled by governments

and regulators)

Reliable electricity supply is important for networks, and for users to recharge their

devices

Existing networks that can be shared, including domestic networks and international

connectivity, are important in minimising roll-out costs.

Policy and

regulation

In addition to competition and spectrum management, this includes economic regulation

of telecoms networks and services. Obligations attached to telecoms licences are

important drivers of roll-out. Also, policy makers increasingly include national broadband

plans as a key policy plank to spur wider economic development. Finally, innovation in

network roll-out is beginning to involve other areas of regulation. For example, Google’s

Project Loon (see Section 4) is increasing the relevance of aeronautical regulation.

Drivers and barriers related to take-up

Cost and

device

availability

Does the cost to the consumer of a fixed or mobile Internet service make it unaffordable?

Is the cost and payment schedule of a device a barrier to affordability? Are attractive

devices available in the market?

These aspects can be influenced by the market structure and strategy of device

manufacturers, whether IAPs subsidise devices, and whether there are restrictive

government policies or taxes (including import restrictions or duties).

Social

factors

Are consumers able to use the Internet, driven by their education and awareness or by

cultural restrictions? This include awareness of the benefits of Internet usage (which is

influenced by both literacy and IT literacy); and whether the use of the Internet is deemed

appropriate for all users in certain communities.

Drivers and barriers related to usage

Means of

access

How do consumers access the Internet? Is it on a PC at home or work, on a smartphone

or tablet, or in an Internet café? This affects the content and applications that a consumer

uses: for example, long videos are easier to consume at home.

Relevant,

accessible

content

Does interesting content exist in the local language, made by local or global CAPs?

Quality of

service

Can the consumer receive content at sufficiently high bandwidths and with low enough

latency? This is important for advanced content.

Trust What is the perceived and actual level of security, driven by actual security and

consumer awareness of security. Is Internet access broadly free and private? Monitoring

and censorship can make consumers feel uncomfortable about accessing content freely.


Ref: 2004569-383

‘Unconnected’ countries have poorer outcomes due to the greater impact of key barriers

‘Unconnected’ countries36

perform worse than the other clusters on all three outcomes discussed

above. As illustrated in Figure 3.22, they have lower availability (as measured by 3G coverage)

and lower take-up (as measured by 3G and 4G penetration). Moreover, users in ‘unconnected’

countries who do take up Internet services also have substantially lower usage than users in other

clusters (as shown in Figure 3.23).

Figure 3.22: 3G coverage and 3G/4G take-up by

cluster [Source: GSMA, Analysys Mason, 2015]

Figure 3.23: Annual data usage per Internet

subscription, GB per year [Source: Analysys Mason,

2015]

These poorer outcomes are in part driven by the greater impact of key barriers. We saw in

Section 3.1 that a lack of awareness is a key barrier to take-up, and this is also the case in

‘unconnected’ countries. This barrier affects the business case for network roll-out as well. In

addition, a number of other barriers apply to a greater extent in ‘unconnected’ countries than in

other clusters:

The cost of service plans and handsets acts as a barrier to take-up

Low PC penetration affects the types of content that users can access at a reasonable quality

There is less relevant content available in local languages.

Here we explore the extent to which these barriers are a more significant factor in the

‘unconnected’ countries. Further details of the extent of the full set of barriers in all three clusters

can be found in Annex C.

36 That is, Papua New Guinea, Bangladesh, Pakistan, Nepal, Myanmar and Laos.

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

3G coverage 3G/4G take up

Connected Connecting Unconnected0

50

100

150

200

Connected Connecting Unconnected


Ref: 2004569-383

As discussed in Section 3.1 earlier, cost was not the primary barrier in ‘connecting’ countries. In

‘unconnected’ countries, however, affordability of service plans does act as a barrier to take-up, as

shown in Figure 3.24 and Figure 3.25 (based on the cheapest entry-level fixed and mobile

packages in the three clusters relative to gross national income).37

Figure 3.24: Price of a 500MB prepaid handset-

based mobile broadband package as a % of GNI per

capita [Source: ITU, 2014]

Figure 3.25: Price of an entry-level fixed broadband

package as a % of GNI per capita [Source: ITU,

2014]

PC and smartphone penetration affect how users can access the Internet. The rate of PC

penetration can also suggest how many multi-device Internet users there are. High PC penetration

indicates that many users can access the Internet through their PC, but these users may also have

access through a smartphone. However, low PC penetration means that most users in a country, if

they are accessing the Internet at all, are likely only to be able to do so through a mobile device.

As illustrated in Figure 3.26, PC penetration is low in the ‘unconnected’ cluster, suggesting that a

lack of multiple-device users may constrain usage.

37 The Broadband Commission for Digital Development, a Paris-based body formed by the ITU and UNESCO, views

broadband as affordable if it costs less than 5% of gross national income.

0%

2%

4%

6%

8%

10%

12%

14%


Mobile broadband Affordable

0%

2%

4%

6%

8%

10%

12%

14%


Fixed broadband Affordable


Ref: 2004569-383

Figure 3.26: PC penetration of households [Source: Analysys Mason, Euromonitor, 2015]

Finally, a lack of relevant content may be a more significant barrier in ‘unconnected’ countries. The

most relevant content is that which is in the local language, either developed by local CAPs or by

international CAPs but optimised for local language and context. It appears that a lack of local-

language content acts as a more significant barrier in ‘unconnected’ countries. This is shown in

Figure 3.27.


of the top-40 websites

that are in local

languages [Source:

Alexa, Analysys Mason,

2015]

CAPs and IAPs are aware of these barriers that exist in the ‘unconnected’ cluster. They are

currently implementing an array of partnerships and programmes to address them. We explore

these more fully in the next section.

0%

20%

40%

60%

80%

100%

200

5

200

6

200

7

200

8

200

9

201

0

201

1

201

2

201

3

201

4

Australia Hong Kong Indonesia Japan

Malaysia New Zealand Pakistan Philippines

Singapore South Korea Taiwan Thailand

Vietnam Connected Connecting Unconnected

0%

10%

20%

30%

40%

50%

60%

70%

Local languageand local CAP

Local languageand international

CAP

Not in locallanguage

Pro

po

rtio

n o

f th

e to

p 4

0 w

eb

site

s



Ref: 2004569-383

4 Partnerships and a forward-looking policy environment are

essential for unlocking further the benefits of the Internet

In this section we explore how the barriers to further broadband growth in Asia–Pacific can be

reduced or removed.

CAPs are becoming essential partners in the development of broadband Internet opportunities.

They are forming partnerships with IAPs and other organisations to help lower the barriers to

Internet adoption and create new opportunities for mutual revenue growth.

Likewise, certain policy makers in the region are implementing innovative programmes that

help to lower barriers to adoption, sometimes in partnership with IAPs or CAPs. This is

helping to drive the adoption of broadband services, particularly in challenging environments

and less-developed countries.

In addition, policy makers and regulators also play an essential role by ensuring that the

environment they influence is as conducive as possible to investment and innovation, by IAPs,

CAPs and others. Sustainable competition is essential, as is a regulatory framework that

provides timely and transparent support for initiatives to accelerate the availability, take-up

and usage of the Internet.

4.1 CAPs and IAPs are forming innovative partnerships to address these barriers

CAPs and IAPs are increasingly working together to address barriers to Internet availability, take-

up and usage. This is to their mutual benefit, as it creates new opportunities for IAPs to sell access

to their networks, and for CAPs to expand their user base. A number of examples of these

partnerships in Asia–Pacific are presented below, in the form of case studies. An overview of the

partnerships covered is provided in Figure 4.1.

Figure 4.1: Overview of example CAP partnerships [Source: Analysys Mason, 2015]

CAP Partnership Barrier addressed Outcome / beneficiaries

Partnerships that address availability

O3B /

Google

Satellite mobile backhaul Business case for

network roll-out

Lower roll-out costs for IAPs

Better availability for end users

Google Project Loon Business case for

network roll-out



Google Titan Aerospace Business case for

network roll-out



Facebook Satellite broadband

access coverage in

partnership with Eutelsat

Business case for

network roll-out



Facebook Project Aquila Business case for Lower roll-out costs for IAPs


Ref: 2004569-383

CAP Partnership Barrier addressed Outcome / beneficiaries

network roll-out Better availability for end users

Partnerships that address take-up

Mozilla Low-cost smartphones Device cost and

availability

Lower costs for end users

More customers for IAPs and CAPs

Google Android One programme

to develop low-cost

smartphones

Device cost and

availability

Lower costs for end users

More customers for IAPs and CAPs

Facebook Free Basics programme

to provide limited zero-

rated Internet access

Social factors,

including

awareness

Greater awareness and lower costs

for end users

More take-up for IAPs

Partnerships that address usage

Mozilla Mozilla handset

localisation programme

Relevant content in

local languages

More relevant content for users

Greater usage for CAPs

Google Android carrier billing Relevant content Easier access to relevant content

for users

More usage revenue for CAPs and

IAPs

Facebook Partnership with

Ericsson and Axiata to

improve quality of

service in Indonesia

Quality of service Better experience for users

More usage revenue for CAPs and

IAPs

We discuss the partnerships relevant to each outcome below. For each initiative, we examine the

nature of the barrier addressed, how this has been achieved (including a look at some of the key

enablers), and the desired outcome (including an assessment of who the main beneficiaries are).

IAPs and CAPs are working together to lower the cost of network availability

The partnerships we have identified that lower barriers to availability focus on improving the

business case for network roll-out, especially in challenging, high-cost locations.

Google’s Project Loon is a very new and innovative example of CAPs partnering with IAPs to

improve availability. It focusses on lowering the cost of network roll-out in very remote areas,

improving the IAPs’ business case. Project Loon uses balloons that travel at high altitudes to

provide high-speed wireless broadband in partnership with local carriers.

Project Loon relies on a number of enablers for its success. First, it requires an IAP partner. This is

typically a mobile operator who provides the necessary radio spectrum, and owns the relationship

with the end customer. It also relies on a conducive regulatory environment, in particular the

ability to use radio spectrum to communicate with the balloons. Aeronautical regulation is also

relevant, since a large network of balloons will need to coordinate with other users of the airspace

regularly.

Project Loon is currently in its early stages, with plans to test it further in the Asia-Pacific region

(in Sri Lanka, for example). However, if successful, it will benefit end users, who will gain better


Ref: 2004569-383

mobile coverage, IAPs, whose business case will be improved, and CAPs, who will be able to

address a larger population of Internet users.

In addition to Project Loon, Google has acquired Titan Aerospace, which aims to provide Internet

access through light-weight solar-powered aircraft. These can provide bandwidth to a targeted area

on the ground, complementing Project Loon: the balloons could provide wide area coverage, and

the Titan aircraft could provide extra capacity to localised areas.

Titan relies on a number of factors for its success. It relies on IAP partners who provide the service

to end customers on the ground. It also relies on a conducive regulatory environment, in a similar

way to Project Loon.

Titan would also benefit end users in remote areas, who can gain extra Internet capacity when

needed. This is especially the case in natural disaster areas, such as areas hit by earthquakes, as

these have been identified as a key target area for the initiative.

O3B (backed by Google and a number of financial investors) is also partnering with IAPs to

improve availability. It also focuses on lowering barriers to a successful business plan. It does this

by reducing the cost of backhaul links from mobile towers. O3B uses low-Earth-orbit satellites (at

an altitude of 8000km) to provide mobile backhaul to IAPs.

For its success, it relies on demand from and partnerships with mobile operators. It also requires

radio spectrum to operate, although O3B acquires the licences for this itself rather than relying on

an IAP partner.

Like Project Loon, if successful O3B would bring benefits to end users, IAPs and CAPs. O3B has

reportedly already struck a number of partnerships in the region, including in Papua New Guinea,

Pakistan and Brunei Darussalam.

Facebook’s partnership with Eutelsat to offer satellite broadband access services to end users in

Sub-Saharan Africa. This project uses the Ka-band AMOS-6 satellite, and also provides gateways

and terminals on the ground.

It is not clear, at the time of writing, what the exact business model will be. However, in all cases

the success of this partnership will rely on access to satellite capacity, and the smooth distribution

of terminals to on the ground customers in potentially difficult to reach places.

Like Project Loon, if successful this partnership would bring benefits to end users, IAPs and

CAPs. These benefits would be felt in Sub-Saharan Africa, but the project could also provide a

model for remote parts of Asia.

Finally, Facebook’s Aquila unmanned aerial vehicle programme, which could work in a similar

way as the Titan Aerospace programme (see above), envisages solar powered drones to provide

Internet access services to customers on the ground. The drones will communicate with end users

using free space optical signals.


Ref: 2004569-383

Project Aquila is expected to rely on IAPs to service the end customers on the ground as it

provides capacity to these IAPs on a wholesale basis. Its success will be dependent on enabling

regulation being in place (including both in the telecoms and Aerospace areas), given the need to

coordinate with other users of airspace.

Beneficiaries of Project Aquila would include end users who would receive better coverage, IAPs

who would gain access to more customers at a lower cost, and CAPs, who would have access to

more end customers for their content and applications.

CAPs are developing partnerships that encourage non-adopters to take up broadband packages

CAPs are entering into partnerships that address barriers to take-up. These seek to address the

issues of cost and availability of handsets, and social factors related to awareness of the usefulness

of the Internet.

Mozilla, the CAP behind the Firefox Internet browser, is working to lower barriers to smartphone

affordability. It is doing this by developing a smartphone based on its Firefox open-source

operating system.

Mozilla’s core partners are device manufacturers such as Cherry Mobile in the Philippines and

Symphony in Bangladesh to manufacture its devices. It also relies, in part, on IAPs to sell and

distribute handsets to end users, and to include them in service plans. Finally, given the low cost

nature of the smart phones, low taxation on devices is crucial.

The key beneficiaries of this partnership are price-sensitive consumers in developing markets.

IAPs also benefit, as the project helps to extend mobile penetration to new market segments.

Finally, CAPs benefit from having a larger population of Internet users who can access their

content and applications.

Google is also working to lower barriers to smartphone affordability through its Android One

project. It is working with regional manufacturers to develop an Android handset targeted at cost-

conscious consumers in the developing world.

Manufacturing partners include Cherry Mobile, Spice, Symphony, Karbonn and Micromax. As for

Firefox Mobile, success for Android One will also rely in part on IAPs to distribute the devices. Its

success is also sensitive to the taxation attracted by mobile devices.

The beneficiaries of this project are also price-sensitive consumers in developing markets,

including those in Indonesia, Bangladesh, Myanmar, Nepal, Sri Lanka and Pakistan where the

project has been launched. IAPs and CAPs also benefit, as the population of potential customers

increases in these markets.

Facebook is working to improve awareness of the benefits of the Internet in developing markets

through its Free Basics initiative. It is working with IAPs to provide “zero-rated” access to certain

Internet content (e.g. Wikipedia, Facebook, health and medical sites and local news sites). The


Ref: 2004569-383

hope is that, once price-sensitive users have experienced a limited selection of Internet content,

they will be more willing to take up a full Internet access service.

Free Basics is implemented in partnership with IAPs, and relies on them for its success. The

incentive for IAPs is that in return of forgoing usage revenue from a certain set of customers in the

short term, they will gain future revenue from greater take up of full service plans.

Key beneficiaries of Free Basics are the end users themselves, who gain low-cost access to certain

content. These users also benefit if they decide to take up a full service plan. This behaviour also

benefits IAPs. Finally, Facebook benefits by signing up new users to its service. In Asia–Pacific,

Free Basics has launched in the Philippines, Indonesia, and India.

CAPs and IAPs are working to drive usage among under-served customers

CAPs and IAPs are forming partnerships that address barriers to usage. In particular, they are

targeting barriers to the development and usage of relevant content, and barriers associated with

the quality of the user experience.

Mozilla is addressing barriers to relevant content in local languages through its Localisation

Project. This involves translating its Firefox operating system into over 90 languages, and

encouraging local-language content developers to distribute content through the Firefox

Marketplace.

This project relies on two sets of partners for its success. First, IAPs must distribute Firefox

handsets. In addition, it is important that local developers create enough content for the Firefox

Marketplace to make it successful. For example, Firefox-based applications such as OLAcabs and

Bikroy are popular in Bangladesh, and Easy Taxis and OLX classifieds are popular in the

Philippines.38

This project potentially benefits users who speak languages that are poorly served by the Internet.

It also benefits the CAPs which develop apps for the Firefox marketplace, and IAPs which gain

new customers due to the desire to access local content.

Google is addressing barriers to the payment for relevant content and applications through its

carrier billing partnerships. These integrate the Google Play store with IAP billing systems,

allowing customers to purchase apps and content and have the cost charged to their mobile

account. It helps users who have no bank account by removing the requirement for a credit card,

and drives usage among the wider user base by increasing the convenience of content purchases.

The project relies on IAP partners as key enablers. It involves substantial IT integration work to

connect IAP billing systems with the Google Play store, which requires a high level of trust from

the IAPs. It also shifts the responsibility for credit checks and cash management to the IAPs.

38 See https://blog.mozilla.org/blog/2014/12/16/firefox-os-expands-to-nearly-30-countries/.


Ref: 2004569-383

This partnership has wide benefits: end users gain easier access to a wide range of content; IAPs

can differentiate their services and gain more revenues from higher data usage; and CAPs can sell

more applications and content. These benefits have been experienced across the region, with

partnerships set up in Australia, Hong Kong, Indonesia, Japan, South Korea, New Zealand, the

Philippines, Singapore, Taiwan and Thailand.

Facebook is partnering to improve the quality of service experienced by its end users in Indonesia.

It is working with Ericsson and XL Axiata to develop a new methodology to improve quality of

service, especially for those customers who access the Internet on a 2G connection.

The project is enabled by a close working partnership among the three entities involved. In

particular, it relies on XL Axiata giving access to its network to engineers from Facebook. This

means it is a deep partnership that relies on significant trust between the parties.

If the project is implemented throughout the XL Axiata network, it will benefit customers by

giving them a better experience of the Internet. It will also benefit XL Axiata, which will be able

to differentiate its service due to the quality improvement. Finally, it will drive Facebook usage.

4.2 Some policy makers are implementing innovative policy initiatives to lower barriers,

often in partnership with IAPs or CAPs

As discussed above, policy makers are also working to address barriers to Internet availability,

take-up and usage, often in partnership with IAPs or CAPs. In this section we again take a case-

study based approach to examining these policy initiatives, presenting a number of examples

targeted at each of the three outcomes. For each initiative, we examine the nature of the barrier

addressed, how this has been achieved, and the outcome (including an assessment of who the main

beneficiaries are). An overview of the policy initiatives covered is shown in Figure 4.2.

Figure 4.2: Overview of example policy initiatives [Source: Analysys Mason, 2015]


Initiatives that address availability

Japan Public funding for

broadband roll-out

Business case for

network roll-out



users

South Korea Public funding for

broadband roll-out

Business case for

network roll-out



users

Singapore Development of national

broadband network

Business case for

network roll-out



users

Initiatives that address take-up

South Korea ICT education programme Social factors,

including awareness

of the benefits of the

Internet

Improved awareness for

~16 million people,

potentially improving take-

up and usage

Malaysia 1Malaysia netbook Affordability of PC penetration increased


Ref: 2004569-383


distribution programme connected devices to ~65%, lower a key

barrier for users to take up

broadband

Initiatives that address usage

Japan Investments in e-

government

Relevant

accessible content



users

Malaysia Investments in

e-government, e-education

and e-health

Relevant

accessible content



users

These initiatives tend to blend fixed and mobile broadband access, recognising the different roles

that each technology plays in different contexts. For example, although Singapore has been at the

forefront of fixed fibre broadband rollout, it is also investing significantly in ensuring its wireless

infrastructure remains world-class. Examples include timely and proactive release of spectrum for

mobile use, as well as a world-leading position in the field of TV White Spaces (TVWS), where it

has shown leadership by hosting detailed trials and issuing usage rules alongside the USA and the

UK.

We discuss each of these initiatives in more detail below.

Government financial support can lower barriers to availability when it is transparent and

enhances competition

Governments in Asia–Pacific have made public investment in advanced broadband access and

core networks. These programmes have addressed barriers to a successful business case for

network construction, by contributing to the cost of large-scale access network roll-outs. In

Singapore, Japan and South Korea, this has helped to create some of the most advanced networks

in the world. As a result, this has benefited end users, as well as IAPs which have received funding

or gained access to advanced networks at regulated wholesale rates, and attracted extra customers.

In all of these cases, investments have been transparent and have promoted competition. In South

Korea and Japan, financing was provided to all IAPs. In Singapore, the new network is open

access at both the retail and wholesale levels.

An overview of these three programmes is set out below.

South Korea:

public investment

in fibre access

networks

Beginning in 1999, South Korea made investments in fibre access networks

in partnership with IAPs. Between 1999 and 2005, the government made

USD 1.8 billion in low-cost loans available to all fixed IAPs. In partnership

with this, the private sector invested USD 14.5 billion over the same period.

In addition, the government has also invested USD24 billion in the

construction of a national backbone network.

Japan:

public investment

From 2000, the government of Japan provided financing to IAPs for fibre

broadband by offering zero-interest loans, and by providing debt guarantees.


Ref: 2004569-383

in fibre access

networks

The government also provided tax incentives, by allowing IAPs to

depreciate one third of the cost of these fibre broadband investments in the

first year of deployment. Funding was available to all IAPs which were

prepared to invest.

Singapore:

investment in

national broadband

network

In 2006, Singapore announced plans to invest in a nationwide fibre-to-the-

home (FTTH) network, known as the NGNBN. At the time, the two main

fixed IAPs, SingNet (with a DSL network) and StarHub (with cable) had

near-ubiquitous coverage. The network was structured to be open access,

with a regulated operator (NetCo) providing duct and dark fibre to a number

of operating companies (OpCos) who install active electronics and provide

wholesale services to a number of retail service providers (RSPs).

Figure 4.3 shows the impact of these investments. Since the launch of publicly supported fibre

networks, take-up of fibre broadband services has increased rapidly, with the majority of Japanese

and South Korean subscribers on a fibre connection, and close to half of Singaporean subscribers.


of fixed broadband

subscribers in Japan,

Singapore and South

Korea with a fibre

connection [Source:

TeleGeography, 2015]

A number of large-scale programmes and partnerships have been successful in lowering barriers

to take-up in the region

The governments in South Korea and Malaysia have had some success in lowering barriers to

take-up through large-scale policy programmes. In South Korea, an education programme helped

to lower the social barriers to take-up, by improving awareness of the benefits of the Internet

among under-served social groups. The key beneficiaries of this were the Internet users

themselves, but also IAPs and CAPs which received new customers as a result of improved

awareness. In Malaysia, the government, in partnership with major device manufacturers, has

increased PC penetration to levels seen in much more advanced countries. This has benefited the

0%

10%

20%

30%

40%

50%

60%

70%

2006 2007 2008 2009 2010 2011 2012 2013 2014

Pro

po

rtio

n o

f fixe

d b

roa

db

an

d s

ub

scri

be

rs o

n a

fib

re c

on

ne

ctio

n

Japan Singapore South Korea


Ref: 2004569-383

end users who received computers, as well as IAPs and CAPs which attracted new customers as a

result of the lowering of this barrier. An overview of these programmes is set out below.

South Korea:

IT education

programme

In South Korea, the government implemented a series of national IT training

programmes. The first was organised for 10 million people in 2000. In

addition to this, IT literacy programmes were offered to marginalised groups

such as low-income households, farmers, fishermen, housewives, the

disabled, the elderly and prisoners. 5.8 million people attended these

programmes over 2000 and 2001. Tailored educational tools were used to

reach the various groups, involving a combination of distance learning and

house calls for those less able to leave home.

Malaysia:

1Malaysia netbook

distribution

programme

The Malaysian government implemented a large-scale programme to

increase PC penetration in remote and rural areas. The 1Malaysia netbook

distribution programme provided 1.6 million laptops free of charge to

inhabitants in rural areas. For consumers who were not eligible for this, in

the 2008 budget the Malaysian government announced that broadband and

consumer access device equipment was to be exempt from import and sales

taxes. Although there has been some controversy regarding the quality of

the laptops provided, the programme helped to increase PC penetration in

Malaysia from approximately 31% to 67% of households over the five-year

period to 2012, to a level similar to that seen in more-advanced countries.

These programmes have had an impact on South Korean broadband penetration, helping to drive

take-up, and Malaysian PC penetration, as shown in Figure 4.4.

Figure 4.4: Fixed

broadband penetration

in South Korea [Source:

TeleGeography,

Euromonitor, 2015]

0%

20%

40%

60%

80%

100%

120%

2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014

Pro

po

rtio

n o

f h

ou

se

ho

lds

Korea fixed broadband penetration

Malaysia PC penetration

Start of 1Malaysia programme


Ref: 2004569-383

Governments have sought to lower barriers to usage by encouraging local content creation

The governments in Singapore and Malaysia have taken steps to encourage local content creation

and usage. The Singapore government promotes local content production and distribution, and the

Malaysian government has invested in e-government, e-health and e-education. This is of benefit

to users, who gain access to more relevant, localised content, and to the providers of content

themselves.

Japan’s investment

in e-government

The government of Japan has invested in putting the majority of government

services online, in order to make government processes more efficient and

drive internet usage. This investment happened relatively early in the

development of the Internet. By 2005, 95% of government forms and

notifications were online.

Malaysia:

investment in

e-government,

e-health and

e-education

The Malaysian government has set up a programme to make government,

health and learning services available online. As of 2014, over 70% of

government services were available online. It has set further targets in 2015,

aiming to implement e-health applications in 200 health facilities. The

government has also set targets for greater utilisation of online government

services. Finally, the government is also developing an e-learning portal,

entitled the Virtual Leaning Environment (VLE), which has learning

resources for students, teachers and parents. As part of this programme, the

Ministry of Education has deployed over 100 000 laptops in schools.

4.3 Policy makers have an essential role to play in ensuring that barriers to innovation

and investment are as low as possible

In addition to public interventions in demand and supply, directly and through public-private

partnerships as discussed in Section 4.2, policy makers play a crucial role in lowering Internet

barriers by creating an attractive regulatory environment. Here we discuss how good public policy

practices can help to achieve this. We draw lessons from the outcomes achieved by countries in the

‘connected’, ‘connecting’ and ‘unconnected clusters, and from our discussion of key barriers in

Section 3.339

.

Examples of good public policy practice include:

Sector liberalisation and the encouragement of sustainable competition, including a non-

restrictive licensing regime and spectrum policy that ensures underutilised radio spectrum can

be freed up for innovative use.

39 This builds on work we have conducted for regulators and operators in the region (including the GSMA), and for

NGOs and industry groups such as the Internet Society in Africa and Latin America.


Ref: 2004569-383

Effective regulation that strikes a balance between encouraging competition and stimulating

investment and innovation

Business conditions conducive to private investment locally, including ensuring it is easy to do

business, that taxation on Internet access services and devices is not excessive, and policies

that lower the cost of network construction are in place; and

Carefully balanced data protection and localisation regulations that foster trust and preserve

the fundamental rights of users, whilst encouraging innovation and enabling international

content and applications to be delivered to end-users.

Liberalising the regulatory regime can lower barriers by encouraging competition and investment

Since the development of competitive telecoms markets in the 1990s, the benefits of sector

liberalisation have been well known and understood. Allowing investment by new entrants with

minimal restrictions can encourage competition and investment. However, once the sector has

been officially liberalised, market outcomes are influenced by good policy regarding licensing and

the allocation of scarce resources such as radio spectrum. Best practice regarding the provision of

licences to IAPs reflects the following principles:

Limits to the number of licences available can constrain competition and investment. Except

where justified by the scarcity of resources such as radio spectrum, licences should be easy to

obtain and rights they imply should be easy to exercise.

Providing separate licences for infrastructure and service provision can lower barriers to entry

and encourage competition. For example, even if there are a limited number of submarine

cable landing licences, allowing any licenced operator to access the landing station can

increase competition. This is common practice in many markets, including Singapore for

example, with separate licenses for Facilities Based Operators and Service Based Operators.

The conditions and obligations attached to the licences themselves should not be overly

onerous. For example, they should be, where possible, technology neutral and allow the

licensee to offer a wide range of services. This is key to enabling innovative deployment

strategies, such as those involving aeronautical vehicles and techniques (e.g. balloons, satellite

and drones), as by definition they do not easily fit within existing licence types.

Where possible, exemptions from the requirement to obtain a specific licence can encourage

investment. An example of good practice in this respect is the European Union, which has a

general authorisation regime. Telecommunications operators do not need a licence, but instead

must inform the national regulatory authority of their intention to operate40

. Class licences go

40 Directive 2002/20/EC of the European Parliament and of the Council of 7 March 2002 on the authorisation of

electronic communications networks and services (Authorisation Directive)


Ref: 2004569-383

some way to enabling a light licensing regime, but in many cases it may be possible for

licenses to be dispensed with entirely (for example for online content).

Where resources are currently scarce, regulators should first seek to expand supply, for

example by releasing underutilised spectrum for new uses; in assigning these resources,

regulators should also follow a transparent process that encourages innovation and investment.

Governments should ensure abundant access to both licensed and unlicensed spectrum at high,

medium, and low frequencies. Finally, new models of spectrum sharing that allow dynamic

and efficient use should be encouraged. For example, technologies and practices could be

encouraged that allows spectrum to be shared dynamically with new users when not in use by

existing users.

Effective regulation strikes a balance between encouraging competition and stimulating

investment and innovation

Most consumers in Asia-Pacific are enjoying unprecedented choice for Internet access services,

often at prices that are remarkably low by global standards. Choice and low prices are hallmarks of

successful competition, which regulators and policy-makers must continue to nurture, whilst

ensuring a flexible approach that stimulates and rewards investment in innovative approaches to

broadband connectivity. Competition remedies should:

focus on clearly defined markets where competition issues can be proven to exist.

be proportionate, i.e. a regulator should apply the least burdensome remedy that achieves the

desired outcome

be transparent and non-discriminatory, applying to all access seekers equally.

Importantly, best practice regulation also implies forbearance where competition issues are not

clear or are likely to be transitory. In new markets related to the delivery of online content and

applications, market conditions continue to evolve too fast for policy-makers and regulators to risk

intervening in the wrong way, and therefore many advanced regulators have taken a wait-and-see

approach, arguably to the benefit of innovation and consumers.

This is particularly relevant both to IAPs, who continually need to invest in expanding and

improving their networks, and to CAPs who operate the platform and services that attract end-

users to the Internet.

Platforms are a major enablers of content discovery and demand aggregation online. Accordingly,

they have an important role to play in reducing the difficulty and costs of rolling out relevant

content to end-users, and enabling the monetisation of this content, often in partnerships with IAPs

(e.g. through carrier billing).

Effective regulation should seek to encourage, and not deter or hinder, the rollout of networks in

more risky areas and of large-scale platforms to the benefit of end-users and of local CAPs,


Ref: 2004569-383

particularly in ‘unconnected’ and ‘connecting’ countries, where they can help bring relevant and

accessible content to drive adoption of broadband services.

Providing an attractive investment environment can encourage long-term investment

Given the long-term and capital-intensive nature of investments in broadband infrastructure, a

local environment that is attractive for private investment is linked to positive outcomes in the

broadband market. There are four main aspects to this.

The overall attractiveness of the investment environment, including the barriers to foreign

direct investment, the ease of starting a business, enforcing contracts, obtaining permits, and

other key business inputs;

The level of sector-specific taxation, which, if excessive, can distort investment decisions and

discourage take-up and usage;

The level of censorship and the monitoring of end users, which should be kept to a minimum;

and

Policies that lower barriers to the construction of fibre access, backhaul, and backbone

networks.

There is wide variation in the attractiveness of the investment environments of the countries in

Asia–Pacific. The World Bank publishes an index that ranks countries according to their ease of

doing business, which has a major impact on foreign direct investment. This covers metrics such

as how easy and transparent it is to start a business, obtain a construction permit, pay tax, enforce

contracts and resolve insolvency. ‘Connected’ countries in Asia–Pacific consistently achieve a

high ranking in this index (as shown in Figure 4.5).

Cluster Average rank, World Bank ease

of doing business index

Figure 4.5: Average rank,

World Bank ease of doing

business index [Source:

World Bank, 2015]

Connected 8.33

Connecting 83.25

Unconnected 144.5

Indonesia is noteworthy in this respect. The government maintains a “negative list” of industries

that are subject to some form of control on foreign direct investment. In many cases, foreign

ownership is restricted, or must be obtained in conjunction with a local partner. Currently,

telecommunications network providers are limited to 65% foreign shareholding.41

The level of taxation on the sector is also important. Whilst there is a trade-off between

government revenue-raising and reduced investment, take-up and usage, the burden of taxation

41 Source: Baker & McKenzie, Client Alert, May 2014, Indonesia Foreign Investment, New Negative List


Ref: 2004569-383

that falls on the broadband sector should not be greater than elsewhere in the economy. Where

possible, lowering the overall burden will encourage investment and take-up.

The GSMA, with Deloitte, publishes research on the level of taxation attracted by mobile devices,

as a proportion of the cost of the device, as illustrated in Figure 4.6. There is a link between the

level of broadband development and lower levels of taxation, with better outcomes associated with

lower tax rates.

Figure 4.6: Taxes

attracted by mobile

devices as a proportion

of handset cost

[Source: GSMA,

Deloitte, 2015]

In addition, investment in the Internet thrives when censorship and the monitoring of end users is

kept to a minimum. Freedom house publishes the Freedom of the Net Index (shown in Figure 4.7)

which measures, amongst other things, limits on content including technical blocking, regulation

and self-censorship. It also measures violations of users rights, such as monitoring and

surveillance. Generally, ‘connected’ countries are freer than ‘unconnected’ countries.

Figure 4.7: Freedom of the Net Index rating of study countries [Source: Freedom of the Net Index, 2014]

Freedom of the Net Connected Connecting Unconnected

Free Australia

Japan

Philippines

Partly free Singapore

South Korea

Cambodia

Indonesia

Malaysia

Sri Lanka

Bangladesh

Indonesia

Not free Thailand

Vietnam

Pakistan

Finally, policies should be in place that lower barriers to the construction of fibre networks. This

includes access networks, but also backhaul networks required to connect mobile base stations and

0% 10% 20% 30% 40%

VN

MY

TH

AU

ID

KH

PH

BD

NZ

NP

MM

PK

LK

Handset taxes as % handset costs



Ref: 2004569-383

fixed network nodes, and the long distance backbone networks that carry Internet traffic around

the world. The following policies can help to achieve this:

Sharing of passive infrastructure, such as poles and ducts (owned by telecoms operators,

utilities, or other parties) should be encouraged as much as possible. This could be through

regulation or other incentives.

Regulations and practices regarding access to non-telecoms infrastructure, such as roads (for

trenching) should be balanced and transparent. A completely open and low-cost approach may

result in disruption due to repeated digging. More appropriate is a transparent, non-

discriminatory approach, which treats all users (e.g. utilities, telecoms operators) equally.

Likewise, conditions that maximise sharing of trenches and co-ordination of dig times can

minimise cost and disruption. For example, a “dig once” policy, which requires the installation

of a single oversized duct which can then be shared, can reduce roll-out costs

Construction permission processes should be streamlined and made as efficient as possible.

For example, if both national and local permissions are required, these could be consolidated

into a single “one-stop shop” decision making processes.

Policy should facilitate interconnection and international connectivity. Regulatory and

licensing restrictions to the construction of new submarine cable landing stations should be

removed to encourage new investments. Likewise, there should be minimal restrictions to the

establishment of new Internet Exchanges.

Data-protection and data-localisation policy that balances innovation, trust and the rights of users

can encourage usage and investment

A final area of policy that is highly relevant to the development of a dynamic, healthy and content-

rich Internet ecosystem relates to what happens to online content and in particular data.

Data-protection and data-localisation policy, whilst providing certain benefits related to trust and

privacy, needs to be carefully traded off against the encouragement of innovation. Data-protection

policy should aim to balance trust and the fundamental rights of users. It should also seek to enable

innovation at as low a compliance cost as possible.

In particular, certain data-protection regulations or restrictions may, while going some way to

establish trust, have a limiting impact on innovation and investment. It is important for

governments to be realistic about this as they balance their competing policy aims.

Requirements that strictly limit the purpose for which data is used may inhibit innovation. This is

because finding hidden patterns in data, and developing uses for these, is a fundamental source of

CAP innovation. This often involves the combination of databases gathered for various purposes,

which may well be different from that of the new service being developed.


Ref: 2004569-383

Limitations on the transfer of data between firms or internationally may also inhibit innovation.

Often, new services are created by combining technical capabilities or databases from different

firms, which may also be in different countries. This type of innovation can be harmed by data-

transfer limitations.

Data-localisation policies, such as those that require certain data to be held in-country, are often

defended on the basis of privacy and the need to ensure local data protection policies are

enforceable. Although in some cases this may be a desirable goal in principle, in practice data

localisation policies have significant downsides:

They can have negative impacts on consumers and the ecosystem. CAPs do naturally seek to

cache their content and data close to end users, as this improves the quality of service. Indeed,

where regulatory and investment conditions and infrastructure are conducive, this process can

happen organically as CAPs optimise their networks.

They are difficult to implement, and even more so to mandate. For these policies to be at all

possible, sufficient underlying infrastructure must be present. For example, data centres used

by CAPs need reliable and low-cost access to power, and access to competitive providers of

domestic and international connectivity. If this is not the case, data-localisation requirements

may prevent CAPs from offering their services. This will be to the detriment of consumers,

who will have a lower choice of innovative services.

They can have unintended consequences that undermine the very rationale for them.

Distributed storage of data in numerous data centres can be more secure than local storage.

This is because of the additional redundancy, and potentially better security of larger,

professional, centralised data centres.

In emerging economies where many industries are developing concurrently, strong data

localisation policies can discourage investment by businesses in many industries, including

banking, financial services, manufacturing, and retailing, who store data globally. If these

businesses are forced to store data locally, it is likely to increase their costs and decrease their

investment.

A more-effective way of achieving data-localisation aims is to ensure that local market conditions

are attractive for investment. This means providing high-quality, low-cost infrastructure, and also

following the good practices outlined in this section.


Ref: 2004569-383

5 Conclusion

The research presented in this report demonstrates clearly, in our view, the impact of the growth of

advanced broadband in Asia-Pacific has had on IAPs and the contributions made by CAPs.

In particular, the tremendous growth of broadband take-up and usage in the region has helped to

create opportunities for profitable investments for IAPs. They have made investments in advanced

networks in order to serve this demand. This has been accompanied by significant top line revenue

growth, a rapid increase in the importance of data and internet services to both revenues and

profits, and healthy returns on the investments made. In addition, new opportunities to invest in

innovative OTT content and services have been unlocked.

CAPs, for their part, are a key driver of growth in broadband take-up and usage. Consumers adopt

broadband access services specifically to use the content and applications produced by CAPs. In

addition, the use of advanced content drives consumers to upgrade their broadband services.

Finally, they are making investments in the physical infrastructure of the Internet that improve the

quality of user experience and save costs for IAPs.

Finally, partnerships between CAPs, IAPs, and other parties, are helping to lower barriers to

Internet development and drive further growth in the region. This, in conjunction with a conducive

policy environment and well placed specific policy interventions can drive future broadband

growth in the region.



Annex A Primary research programme

A.1 Introduction

To gain an in-depth understanding of consumers’ attitudes, and drivers and barriers related to

Internet adoption and usage, we commissioned two end-user surveys in several Asian markets in

August 2015:

‘Mobile survey’: a survey of mobile users in five markets (Australia, Indonesia, Singapore,

Sri Lanka and Thailand) conducted on survey participants’ mobile telephones (smartphones

and higher-end feature phones). All participants in this survey were mobile Internet users, and

a proportion of them also subscribed to a fixed Internet connection.

CATI research programme: computer-assisted telephone42

interviews (CATI) with people in

three emerging Asia–Pacific markets (Indonesia, Sri Lanka and Thailand). All participants in

this survey were full non-adopters of the Internet, i.e. none of them subscribes to either a

mobile Internet service or a fixed Internet service.43

The reason for carrying out these two different types of primary research programme was to reach

both adopters of the Internet44

in the target markets and full non-adopters of the Internet,45

particularly in the three emerging markets46

covered by our primary research programme.

We achieved a robust sample size of more than 5000 people. Figure A.1 provides detailed

information on the sample sizes by country and survey type:

Figure A.1: Mobile and CATI survey sample sizes [Source: Analysys Mason, 2015]

Mobile survey CATI survey Total sample

Australia 1000 0 1000

Indonesia 800 400 1200

Singapore 1000 0 1000

Sri Lanka 500 402 902

Thailand 800 415 1215

42 Although CATI refers to telephone interviewing, some of the interviews in Thailand and Sri Lanka were actually

conducted in person. This is because we wanted to be able to reach both urban and rural non-adopters of the Internet, and rural non-adopters often do not have regular access to landline telephones.

43 Please note that some of the CATI survey participants do use the Internet on a regular basis, but none of them

actually pays to receive any Internet service.

44 We were certain that all participants in the mobile survey were Internet adopters, as in order to be able to complete

the survey a person accesses it on their personal mobile device, via a mobile Internet connection.

45 A sufficiently large sample of full non-adopters could not be achieved via an online survey, since a large segment of

non-adopters are also non-users of the Internet, particularly in rural areas. Telephone and face-to-face interviews were more appropriate for this population.

46 Indonesia, Sri Lanka and Thailand.

A–2 | Broadband in Asia–Pacific: How investment, partnerships and policy are driving a global success story

Ref: 2004569-383

Mobile survey CATI survey Total sample

Total 4100 1217 5317

While our samples were not fully representative of the populations of the five markets included in

the research, each country-level sample comprises a reasonably balanced range of people from

different age groups, genders and socio-economic groups. In the following two sections we

provide more information about this and other important aspects of the two surveys.

A.2 Mobile survey

We conducted an online survey of 4100 people in Australia, Indonesia, Singapore, Sri Lanka and

Thailand in August and September 2015. The survey was specifically designed for mobile

devices,47

although a small number of survey participants48

completed the survey on a desktop

computer. Most survey participants completed the survey on their smartphone and tablet devices,

while 20 participants used their feature-phones.49

The mobile survey consisted of 28 questions, most of which were closed-ended multiple-choice

questions. For three of the questions50

we asked the participants to type in a value.51

Four of the

questions were ‘grid’ questions, which required participants to select an answer from a drop-down

menu for each of the questions. The survey was conducted in local languages. In the case of Sri

Lanka, survey participants were given the choice of completing the survey in either English or

Sinhala.

We surveyed 1000 people in each of Australia and Singapore, 800 in each of Indonesia and

Thailand, and 500 in Sri Lanka. The gender distribution was equal52

in Australia, Singapore,

Indonesia and Thailand, but in Sri Lanka we were unable to recruit more than 20% of women.

We only surveyed people aged 18 years and above. Most of the survey participants were between

18 and 35 years old, with around 25% of the sample above 35 years of age. As expected, it was

particularly difficult to recruit mobile Internet users aged more than 65 years in the target

countries, which is why this population is underrepresented in our sample (see Figure A.2).

47 Analysys Mason commissioned “On-Device Research Ltd” to conduct the survey.

48 Only a total of 14 survey participants in Indonesia.

49 Only advanced feature-phones that can connect to the Internet

50 The questions asked how much the participants paid for their current mobile and/or fixed Internet service

subscription; we also asked non-adopters of the fixed Internet how much the service would have to cost in order for them to consider subscribing to it.

51 To ensure that the values typed in by participants made sense, we limited the range of answers which would be

accepted. Furthermore, the validity of this data was later cross-checked against responses to the question regarding the total monthly household expenditure. Around 5% of all entries were excluded from further analysis after this cross-checking exercise. We replaced the excluded entries with newly completed responses and rebalanced the sample.

52 That is, 50% men and 50% women.


Ref: 2004569-383

Figure A.2: Age

distribution of the

mobile survey

participants [Source:


Most of the mobile survey participants came from urban environments. This was expected, as in

the mobile survey we targeted Internet adopters who were also smartphone, tablet or high-end

feature-phone users; that is, a demographic profile which is more typical of urban environments.

Figure A.3 provides more information on the rural and urban split of the samples.

Figure A.3: Rural and

urban split of the mobile

survey participants

[Source: Analysys

Mason, 2015]

Finally, in the profiling questions of the mobile survey we also asked participants to provide

information on their monthly household expenditure.53

While we did not impose strict quotas on

53 More specifically, we asked: ‘Approximately, how much are your monthly family expenses for daily needs, exclude

expenses for recreation, clothes, shoes, and credit payment (e.g. home, car, electrical things, etc.)’. We opted to ask about the monthly household expenditure levels rather than the income levels of the participants, as consumption is

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Over 65

51-65

36-50

26-35

18-25

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

I don't know

Rural area

City


Ref: 2004569-383

various expenditure levels when recruiting survey participants, we asked this question to ensure

that we reached at least some people at both ends of the socio-economic spectrum. Furthermore,

we used this information to check the validity of responses to the open-ended questions about total

spend on Internet services.54

A.3 CATI research programme

In addition to the mobile survey described above, we also conducted a CATI55

primary research

programme with more than 1200 people in August 2015. The main objective of the CATI research

was to understand the key barriers to adoption of the Internet, which is why we exclusively

interviewed non-adopters of the Internet56

across the three markets. The share of non-adopters of

the Internet is fairly small in Australia and Singapore, which is why we limited the CATI research

to the three emerging markets covered by this study – Indonesia, Sri Lanka and Thailand.

The interviews were conducted in interviewees’ native language, either by telephone or in person.

The interview questionnaire consisted of 20 questions. 18 questions were closed-ended, and 2 were

open-ended, with interviewees being asked to state the approximate hypothetical price at which

they would consider subscribing to mobile and fixed Internet services. In addition, we asked

participants who selected the answer “other” to three multiple-choice questions to explain what

was meant by “other”.

Within the CATI survey we interviewed 415 non-adopters of the Internet in Thailand, 402 in Sri

Lanka and 400 in Indonesia. The gender distribution was fairly balanced in Indonesia and Sri

Lanka, and skewed towards males in Thailand.57

Please see Figure A.4 for more information on

the gender distribution of the CATI samples.

usually the preferred welfare indicator, particularly in developing markets. The World Bank states that “consumption is conventionally viewed as the preferred welfare indicator, for practical reasons of reliability and because consumption is thought to better capture long-run welfare levels than current income.” For further discussion of this point, see http://documents.worldbank.org/curated/en/2000/09/17408018/world-development-report-20002001-attacking-poverty.

54 As explained in footnote 49.

55 Analysys Mason commissioned “2-morrow” (Vienna, Austria) to conduct primary research in Indonesia, Sri Lanka

and Thailand, using both computer-assisted telephone interviewing and face-to-face interviews. 2-morrow conducted the field research in cooperation with three local primary research partners.

56 We interviewed people who do not pay for either a fixed or mobile Internet service. Some of these people still use

the Internet at work, school, university etc. but none subscribes to an Internet service plan.

57 Please note that we did not impose any quota on gender, as recruiting full non-adopters of the Internet who are at

least 18 years old was already a challenging task. In fact, in the case of Thailand, our local partner had to contact over 3000 people in order to recruit 400 full non-adopters of the Internet.


Ref: 2004569-383

Figure A.4: Gender

distribution, CATI

[Source: Analysys

Mason, 2015]

We only interviewed people who were aged 18 years or more. Most of the interviewees were

between 26 and 50 in Indonesia and Sri Lanka, and between 36 and 65 in Thailand. See Figure A.5

for more information on the age distribution of the CATI samples.

Figure A.5: Age

distribution, CATI

[Source: Analysys

Mason, 2015]

As mentioned in the previous section, the rural and urban split of the samples was more balanced

in the CATI research programme than in the mobile survey. In the case of CATI research, we

imposed a minimum 30% quota on both rural and urban categories in the samples. Similar to the

situation for gender and age distribution, Thailand’s sample was again somewhat skewed, this time

in favour of participants from an urban environment (see Figure A.6).

51% 48%38%

49% 52%62%

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%


Male Female

17% 15%5%

25% 26%

13%

40% 43%

44%

16% 17%

31%

2% 0% 6%

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%


18-25 26-35 36-50 51-65 Over 65


Ref: 2004569-383

Figure A.6: Rural and

urban split of the

samples, CATI58

[Source: Analysys

Mason, 2015]

As with the mobile survey, we asked interviewees about their monthly household expenditure

levels, to ensure that we interviewed people from the full range of socio-economic groups.

Interviewees were asked to select one of seven ranges of expenditure level: about 50% of samples

in Thailand and Indonesia and 70% of the Sri Lankan sample fell into one of the three options that

lay in the middle of the overall range.59

58 While we gave an option for the participants to say that they did not know whether they lived in rural or urban

environments, none of the CATI interviewees selected this option. It is likely that the interviewers assisted the interviewees who may have been in doubt, an aid which was of course not available to the mobile survey participants.

59 Please note that we did not impose quotas on expenditure levels.

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%


Rural area

City

Broadband in Asia–Pacific: How investment, partnerships and policy are driving a global success story | B–1


Annex B State of the broadband ecosystem in Asia–Pacific

The Asia–Pacific broadband ecosystem is characterised by rapid growth and variation between

countries. In this annex we provide further details of the broadband ecosystem in the 21 countries

of interest to this report.

B.1 The Asia–Pacific Internet picture is one of high growth and wide variations

The Asia–Pacific region makes up a significant part of the world’s population and economy, and is

hugely diverse. In this section we provide further information on the region’s economic and

market performance.

Asia–Pacific has enjoyed rapid, yet uneven, economic growth over the last ten years

The 21 countries studied here carry significant weight in the world economy, representing 18% of

the world’s population and 16% of global GDP.60

Economic and population growth have been

relatively rapid in the last ten years (see Figure B.1 and Figure B.2). Real economic growth rates

averaged 4% per annum (excluding Japan, which drags the overall average down to 2%) and

population growth averaged 1.3% per annum (also excluding Japan).

Figure B.1: Real GDP for Asia–Pacific, constant

2005 US dollars [Source: EIU, 2015]

Figure B.2: Total population, Asia–Pacific [Source:

EIU, 2015]

60 This is equivalent to half of the world’s population and 29% of the world’s GDP, if we include India and China.

0

1

2

3

4

5

6

7

8

9

10

200

5

2006

200

7

200

8

2009

201

0

201

1

201

2

201

3

2014

US

D T

rilli

on

CAGR = 2%

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

200

5

200

6

200

7

2008

200

9

201

0

201

1

2012

201

3

201

4

Bill

ion

CAGR = 1.1%

B–2 | Broadband in Asia–Pacific: How investment, partnerships and policy are driving a global success story

Ref: 2004569-383

The region has established itself as a global leader in broadband Internet

As discussed in Section 2.1, Internet markets in Asia–Pacific have seen strong growth. Take-up of

access services has grown by 28% per annum over the last ten years, and usage of online content

and applications grew by 28% per annum over the last three years.

Although overall data usage has grown more slowly than in some other parts of the world (chiefly

North America and Europe – see Figure B.3), there has been very rapid growth in the use of

international bandwidth for Internet services (see Figure B.4), as the region hosts several of the

most important hubs of the global Internet (in Singapore, Hong Kong and Tokyo).

Figure B.3: Annual data usage (PB) in Asia–Pacific

and worldwide [Source: Analysys Mason, 2015]

Figure B.4: Growth in usage of international Internet

bandwidth by region (indexed to 2010) [Source:

TeleGeography, 2015]

B.2 Three clusters emerge: highly connected world leaders, ambitious broadband

newcomers, and deeply unconnected markets

As discussed in Section 2.1, above, we have classified the 21 markets covered in this report into

three broad clusters: ‘connected’ countries, ‘connecting’ countries and ‘unconnected countries’.

We present further detail on these clusters below, comparing infrastructure, take-up and usage of

broadband services across the clusters.

Infrastructure: ranging from ubiquitous in ‘connected’ countries to poor in ‘unconnected’

countries

The breadth and quality of broadband infrastructure varies significantly between the three clusters.

‘Connected’ countries have ubiquitous 3G coverage and access to very-high-speed fixed

0

100

200

300

400

500

600

700

201

2

201

3

201

4

PB

(0

00

s)

APAC Worldwide

0%

100%

200%

300%

400%

500%

2010 2011 2012 2013 2014Co

mp

ou

nd

gro

wth

sin

ce

20

10

(%

)

Asia + Oceania

Europe

North America

Middle East and Africa

Latin America


Ref: 2004569-383

connections (over 1Gbit/s in countries such as Singapore and Japan). ‘Connecting’ countries also

have good 3G coverage, but have lower fixed broadband speeds available. Although 3G coverage

of ‘unconnected’ countries is improving, they have lower breadth and quality of infrastructure.

This is illustrated in Figure B.5 and Figure B.6.

Figure B.5: 2014 3G coverage as a proportion of

population, by cluster [Source: GSMA, 2015]

Figure B.6: 2014 Maximum fixed broadband speed

available, averaged by cluster [Source:

TeleGeography, Analysys Mason, 2015]

‘Connected’ countries also have far superior international connectivity, as shown in Figure B.7

below.

Figure B.7: International

Internet bandwidth per

Internet user by cluster

[Source:

TeleGeography, World

Bank, 2015]

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%


0

100

200

300

400

500

600

700

800


Mb

it/s

0

20

40

60

80

100

120

140

160

180

200


Kb

it/s

pe

r In

tern

et u

se

r


Ref: 2004569-383

Take-up: ubiquitous in ‘connected’ countries, mobile-dominated in ‘connecting’ countries, and

very low in ‘unconnected’ countries

‘Connected’ countries have higher take-up of both fixed and mobile broadband services. The

‘connecting’ cluster has lower, but growing take-up, reflecting later mobile network roll-outs and

lower fixed broadband speeds. Take-up in ‘unconnected’ countries remains very low. This is

shown in Figure B.8 and Figure B.9. Together with coverage differences discussed above, these

discrepancies in take-up, show that ‘connecting’ and ‘unconnected’ countries suffer from low take-

up even in areas where broadband is available.

Figure B.8: Current 3G and 4G mobile penetration of

population by cluster [Source: GSMA, 2015]

Figure B.9: Current fixed broadband penetration of

households by cluster [Source: TeleGeography,

2015]

Usage: ranging from high usage of sophisticated content to very little usage at all

‘Connected’ countries have higher overall usage than the other two clusters, driven by their higher

fixed take-up, access to high speeds and the availability of more-sophisticated content. Figure B.10

shows annual Internet traffic per capita and Figure B.11 shows the proportion of the top-40

websites that contain sophisticated content.

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%


3G

/ 4

G p

en

etr

atio

n o

f p

op

ula

tio

n

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%


Fix

ed

bro

ad

ba

nd

pe

ne

tra

tio

n o

f h

ou

se

ho

lds


Ref: 2004569-383

Figure B.10: Annual Internet and data traffic (GB)

per capita by cluster [Source: Analysys Mason,

2015]

Figure B.11: Proportion of the top-40 websites that

contain sophisticated content61

[Source: Analysys

Mason, 2015]

This difference in traffic is driven largely by Internet take-up. Looking at the usage per subscriber,

a more-nuanced picture emerges: for fixed users (see Figure B.12), each subscriber consumes

more data in ‘connected’ than in other markets; while for mobile (3G and 4G) subscribers (see

Figure B.13) usage per connection is quite similar in ‘connected’ and ‘connecting’ markets, but

much lower in ‘unconnected markets’, which suggests that international connectivity and access

network capacity are much more constrained in ‘unconnected’ markets.

61 We define sophisticated content as that which needs an advanced broadband connection. It includes video, gaming

and image-heavy websites.

0

50

100

150

200

250

Connected Connecting Unconnected0%

2%

4%

6%

8%

10%

12%

14%

16%


Pro

po

rtio

n o

f to

p 4

0 w

eb

site

s


Ref: 2004569-383

Figure B.12: Annual fixed data traffic (GB) per fixed

broadband subscriber by cluster [Source: Analysys

Mason, 2015]

Figure B.13: Annual mobile data traffic (GB) per

3G/4G mobile connection by cluster [Source:


0

100

200

300

400

500

600


0

10

20

30

40

50

60


Broadband in Asia–Pacific: How investment, partnerships and policy are driving a global success story | C–1


Annex C Key barriers to broadband development in Asia–

Pacific

In this annex we provide further details of the significance of the barriers identified in Section 3.3

in each of the three clusters. We discuss each of the barriers that were not covered in the main

report.

C.1 Availability of broadband requires network deployment to be attractive for

operators, for which policy and regulation play a major enabling role

Here we examine the barriers to availability, and the extent to which they have an impact on Asia–

Pacific markets. The three major barriers identified are:

the business case for network roll-out

the availability of key inputs such as radio spectrum and electric power

the policy and regulatory environment (not covered here, as it was discussed in Section 4.3

earlier).

C.1.1 Business case

The drivers of a successful business case are the cost of inputs, and the country-specific

investment risk. We can compare data that can act as a proxy for each.

The cost of inputs is largely driven by the cost of the access network. This is heavily influenced by

the density of demand. To understand this density of demand, we can examine the number of

Internet users per square kilometre. To examine the country-specific investment risk, we can look

at the index compiled by the World Bank on the ‘ease of doing business’. These are shown in

Figure C.1 and Figure C.2.

C–2 | Broadband in Asia–Pacific: How investment, partnerships and policy are driving a global success story

Ref: 2004569-383

Figure C.1: Internet

users per square

kilometre [Source:

World Bank, 2015]

Cluster Average rank, World Bank ease

of doing business index

Figure C.2: Average rank,

World Bank index of ‘ease

of doing business’

[Source: World Bank,

2015]

Connected 8.33

Connecting 83.25

Unconnected 144.5

We can see from the above that ‘connected’ countries have a significant advantage over countries

in the other clusters in both the density of demand, and the overall business environment.

C.1.2 Availability of key inputs

The key inputs which we have identified as potentially creating barriers to availability are:

the availability of electricity

radio spectrum

the availability of passive network inputs such as duct and telephone poles to investors in

access networks (discussed in Section 4.3).

To some extent, the availability of inputs does appear to be a barrier in ‘connecting’ and

‘unconnected’ countries. Although the level of electrification does not appear to be a barrier (at

above 60% in all three clusters), significantly more radio spectrum has been made available in

‘connected’ countries. These two metrics are shown in Figure C.3 and Figure C.4 below.

0

50

100

150

200

250


Inte

rne

t u

se

rs p

er

sq

ua

re k

m


Ref: 2004569-383

Figure C.3: Electrification as a proportion of

population [Source: World Bank, 2015]

Figure C.4: Average frequency available, MHz per

100 000 population [Source: National regulatory

authorities, TeleGeography, Analysys Mason, 2015]

C.2 Take-up relies partly on affordable devices and services, but also on social factors

and perception

The barriers we have identified to take-up are the cost and availability of devices and service plans

(which we discussed in Section 3.1 and 3.3), and social factors such as education and awareness,

and the appropriateness of Internet services in the local context.

C.2.1 Social factors

Social factors that influence take-up include levels of education and awareness (both general

literacy and awareness of the benefits of the Internet), and the social and cultural acceptance of

accessing Internet content. Awareness is not covered here, as it was discussed in Section 3.1.

Overall literacy rates are high in both ‘connected’ and ‘connecting’ countries, indicating that basic

levels of education are not a major barrier in these countries. In all clusters, more than 60% of the

population is literate (see Figure C.5).

0%

20%

40%

60%

80%

100%


0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

1.8


MH

z p

er

10

0 0

00

po

pu

latio

n

Low frequency High frequency


Ref: 2004569-383

Figure C.5: Average

literacy rate among

adults (aged 15+)

[Source: World Bank,

2015]

In addition, a key factor in take-up is that it must be deemed socially or culturally appropriate or

acceptable to use the Internet. Gender is one example of a social barrier to Internet use. According

to a 2012 study,62

there is a ‘gender gap’ in Internet access. The study estimated that 33% fewer

women in South Asia had access to the Internet than men, while the gender gap was smaller in the

East Asia and Pacific region (at 20%). Worldwide, the study calculated that 23% fewer women

than men are online in developing countries.

C.3 Usage is spurred by trust and the availability and accessibility of relevant content, on

suitable devices and high-quality networks

The barriers we have identified to the usage of Internet content and applications are the means of

access, the availability of relevant content (both of which were discussed in Section 3.3), the

quality of service experienced by end customers, and the level of trust consumers have in the

Internet (both of which are discussed here).

C.3.1 Quality of service

Quality of service (that is, the actual technical performance received by the end user when

consuming content), can have an impact on Internet usage levels. If quality is degraded when

consuming content, an end user will consume less content. Quality of service includes the actual

bandwidth received, as well as other factors such as latency. Akamai, the world’s largest

commercial CDN provider, records actual bandwidths achieved by end users connected to its

network. Whilst it does not collect data for the ‘unconnected’ countries, we can see that users in

62 Women and the web: Bridging the Internet gap, Intel and Dalberg Global Development Advisers, 2012.

0%

20%

40%

60%

80%

100%



Ref: 2004569-383

‘connected’ countries experience higher bandwidths significantly more often than those in

‘connecting’ countries (shown in Figure C.6 for fixed and Figure C.7 for mobile).

Figure C.6: Unique IP addresses connecting to

Akamai from a fixed network at an average speed of

over 4Mbit/s [Source: Akamai, 2015]

Figure C.7: Unique IP addresses connecting to

Akamai from mobile network providers at an average

speed of over 4Mbit/s [Source: Akamai, 2015]

C.3.2 Trust

Internet users will have a higher level of trust in the Internet if they feel that they are unlikely to be

the victims of cybercrime or be exposed to inappropriate content, or are free from fear of penalties

from the government for accessing certain content.

Freedom House, a think tank, produces an annual “Freedom of the Net” index to assess

government censorship and control of the Internet. Of the 14 countries included in our study that

are reviewed by the Freedom of the Net Index, 8 were judged to be on a downward ‘overall

trajectory’, indicated a decline in freedom of the Internet, whilst only 3 were perceived to have

improved their position on this measure between 2013 and 2014.63

The Internet Society’s review

of Asia–Pacific in 2014 highlighted the introduction of stricter Internet legislation in the

Philippines, Vietnam, and Laos, continued ‘heavy Internet oversight’ in Myanmar, and content

filtering and blocking in South Korea, Pakistan, Thailand, Malaysia and Indonesia.64

As

Figure C.8 shows, a large proportion of the countries in this study that appear in the Freedom of

the Net Index are rated ‘partly free’, and three are rated ‘not free’. This implies that censorship

could indeed be a barrier for certain countries in the region, across all three clusters.

63 Freedom of the Net 2014, Freedom House, December 2014.

64 APAC: The Year that Was, The Internet Society, 23 December 2014.

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%


0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%



Ref: 2004569-383

Figure C.8: Freedom of the Net Index rating of study countries [Source: Freedom of the Net Index, 2014]

Freedom of the Net

rating Connected Connecting Unconnected

Free Australia

Japan

Philippines

Partly free Singapore

South Korea

Cambodia

Indonesia

Malaysia

Sri Lanka

Bangladesh

Indonesia

Not free Thailand

Vietnam

Pakistan

broadband in asia€¦ · 0.1 the rise of broadband internet has created opportunities for...

Documents